# Anthropic

## Docs

- [Eine Nachrichtenbatch abbrechen (beta)](https://docs.anthropic.com/de/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/de/api/claude-on-amazon-bedrock): Anthropics Claude-Modelle sind jetzt über Amazon Bedrock allgemein verfügbar.
- [Vertex AI API](https://docs.anthropic.com/de/api/claude-on-vertex-ai): Anthropics Claude-Modelle sind jetzt über [Vertex AI](https://cloud.google.com/vertex-ai) allgemein verfügbar.
- [Client SDKs](https://docs.anthropic.com/de/api/client-sdks): Wir stellen Bibliotheken in Python und TypeScript zur Verfügung, die die Arbeit mit der Anthropic API erleichtern.
- [Create a Text Completion](https://docs.anthropic.com/de/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Erstellen eines Nachrichten-Batches (beta)](https://docs.anthropic.com/de/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Fehler](https://docs.anthropic.com/de/api/errors)
- [Hilfe erhalten](https://docs.anthropic.com/de/api/getting-help): Wir haben versucht, in dieser Dokumentation die Antworten auf die häufigsten Fragen bereitzustellen. Wenn Sie jedoch weitere technische Unterstützung bei der Verwendung von Claude, der Anthropic API oder einem unserer Produkte benötigen, können Sie sich an unser Support-Team unter [support.anthropic.com](https://support.anthropic.com) wenden.
- [Erste Schritte](https://docs.anthropic.com/de/api/getting-started)
- [IP-Adressen](https://docs.anthropic.com/de/api/ip-addresses): Anthropic-Dienste sind unter einem festen Bereich von IP-Adressen erreichbar. Sie können diese zu Ihrer Firewall hinzufügen, um den minimalen Oberflächenbereich für ausgehenden Datenverkehr beim Zugriff auf die Anthropic API und Konsole zu öffnen. Diese Bereiche werden sich nicht ohne Vorankündigung ändern.
- [Nachrichtenstapel auflisten (beta)](https://docs.anthropic.com/de/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/de/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Nachrichtenbatch-Beispiele](https://docs.anthropic.com/de/api/messages-batch-examples): Beispiele für die Verwendung der Message Batches API
- [Nachrichten-Token zählen (Beta)](https://docs.anthropic.com/de/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Nachrichten-Beispiele](https://docs.anthropic.com/de/api/messages-examples): Anfrage- und Antwortbeispiele für die Messages API
- [Streaming von Nachrichten](https://docs.anthropic.com/de/api/messages-streaming)
- [Migration von Text Completions](https://docs.anthropic.com/de/api/migrating-from-text-completions-to-messages): Migration von Text Completions zu Messages
- [Prompt-Validierung](https://docs.anthropic.com/de/api/prompt-validation): Mit Text Completions
- [Nutzungsbeschränkungen](https://docs.anthropic.com/de/api/rate-limits): Um Missbrauch zu verhindern und die Kapazität unserer API zu verwalten, haben wir Beschränkungen für die Nutzung der Claude API durch Organisationen eingeführt.
- [Nachrichtenstapel-Ergebnisse abrufen (beta)](https://docs.anthropic.com/de/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Eine Nachrichtenbatch abrufen (beta)](https://docs.anthropic.com/de/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Streaming von Textvervollständigungen](https://docs.anthropic.com/de/api/streaming)
- [Unterstützte Regionen](https://docs.anthropic.com/de/api/supported-regions): Hier sind die Länder, Regionen und Gebiete, von denen aus wir derzeit den Zugang unterstützen können:
- [Versionen](https://docs.anthropic.com/de/api/versioning): Bei API-Anfragen müssen Sie einen `anthropic-version` Request-Header senden. Zum Beispiel `anthropic-version: 2023-06-01`. Wenn Sie unsere [Client-Bibliotheken](/de/api/client-libraries) verwenden, wird dies automatisch für Sie erledigt.
- [Juli 2024 Updates](https://docs.anthropic.com/de/developer-newsletter/july2024): 28. Juli 2024
- [Überblick](https://docs.anthropic.com/de/developer-newsletter/overview): Entdecken Sie monatliche Updates, technische Einblicke, Best Practices und Erfolgsgeschichten zur Verbesserung Ihrer Claude-Integrationen
- [Modelle](https://docs.anthropic.com/de/docs/about-claude/models): Claude ist eine Familie modernster großer Sprachmodelle, die von Anthropic entwickelt wurden. Dieser Leitfaden stellt unsere Modelle vor und vergleicht ihre Leistung mit älteren Modellen.
- [Sicherheit und Compliance](https://docs.anthropic.com/de/docs/about-claude/security-compliance)
- [Inhaltsmoderation](https://docs.anthropic.com/de/docs/about-claude/use-case-guides/content-moderation): Inhaltsmoderation ist ein wichtiger Aspekt bei der Aufrechterhaltung einer sicheren, respektvollen und produktiven Umgebung in digitalen Anwendungen. In diesem Leitfaden besprechen wir, wie Claude zur Moderation von Inhalten in Ihrer digitalen Anwendung eingesetzt werden kann.
- [Kundenservice-Agent](https://docs.anthropic.com/de/docs/about-claude/use-case-guides/customer-support-chat): Diese Anleitung zeigt, wie man Claudes fortgeschrittene Konversationsfähigkeiten nutzt, um Kundenanfragen in Echtzeit zu bearbeiten, 24/7-Support zu bieten, Wartezeiten zu reduzieren und hohe Support-Volumen mit präzisen Antworten und positiven Interaktionen zu bewältigen.
- [Juristische Zusammenfassung](https://docs.anthropic.com/de/docs/about-claude/use-case-guides/legal-summarization): Diese Anleitung zeigt, wie man Claudes fortschrittliche Verarbeitungsfähigkeiten für natürliche Sprache nutzen kann, um juristische Dokumente effizient zusammenzufassen, wichtige Informationen zu extrahieren und die juristische Recherche zu beschleunigen. Mit Claude können Sie die Überprüfung von Verträgen, die Prozessvorbereitung und regulatorische Arbeiten optimieren, Zeit sparen und die Genauigkeit Ihrer rechtlichen Prozesse sicherstellen.
- [Leitfäden für häufige Anwendungsfälle](https://docs.anthropic.com/de/docs/about-claude/use-case-guides/overview)
- [Ticket-Routing](https://docs.anthropic.com/de/docs/about-claude/use-case-guides/ticket-routing): Diese Anleitung zeigt, wie man Claudes fortgeschrittene Sprachverständnisfähigkeiten nutzt, um Kundenservice-Tickets basierend auf Kundenabsicht, Dringlichkeit, Priorisierung, Kundenprofil und mehr im großen Maßstab zu klassifizieren.
- [Google Sheets Add-on](https://docs.anthropic.com/de/docs/build-with-claude/claude-for-sheets): Die [Claude for Sheets Erweiterung](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) integriert Claude in Google Sheets und ermöglicht es Ihnen, Interaktionen mit Claude direkt in Zellen auszuführen.
- [Computernutzung (Beta)](https://docs.anthropic.com/de/docs/build-with-claude/computer-use): Das aufgewertete Claude 3.5 Sonnet-Modell ist in der Lage, mit [Tools](/de/docs/build-with-claude/tool-use) zu interagieren, die eine Computerdesktopumgebung manipulieren können.
- [Definieren Sie Ihre Erfolgskriterien](https://docs.anthropic.com/de/docs/build-with-claude/define-success)
- [Erstellen Sie starke empirische Evaluierungen](https://docs.anthropic.com/de/docs/build-with-claude/develop-tests)
- [Embeddings](https://docs.anthropic.com/de/docs/build-with-claude/embeddings): Text-Embeddings sind numerische Darstellungen von Text, die es ermöglichen, semantische Ähnlichkeit zu messen. Dieser Leitfaden stellt Embeddings vor, ihre Anwendungen und wie man Embedding-Modelle für Aufgaben wie Suche, Empfehlungen und Anomalieerkennung verwendet.
- [Nachrichten-Batches (Beta)](https://docs.anthropic.com/de/docs/build-with-claude/message-batches)
- [PDF-Unterstützung (Beta)](https://docs.anthropic.com/de/docs/build-with-claude/pdf-support): Die Claude 3.5 Sonnet Modelle unterstützen jetzt PDF-Eingaben und verstehen sowohl Text als auch visuelle Inhalte in Dokumenten.
- [Prompt-Caching (Beta)](https://docs.anthropic.com/de/docs/build-with-claude/prompt-caching)
- [Sei klar, direkt und detailliert](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Lassen Sie Claude nachdenken (Chain of Thought Prompting) um die Leistung zu steigern](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Komplexe Prompts verketten für bessere Leistung](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Tipps für Prompting mit langem Kontext](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Verwenden Sie Beispiele (Multishot-Prompting) um Claudes Verhalten zu steuern](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Überblick über Prompt Engineering](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/overview)
- [Claudes Antwort vorausfüllen für bessere Ausgabekontrolle](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Automatische Erstellung von Prompt-Vorlagen im ersten Entwurf](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Nutzen Sie unseren Prompt-Verbesserer, um Ihre Prompts zu optimieren](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Verwenden von Prompt-Vorlagen und Variablen](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Claude eine Rolle mit einem System-Prompt geben](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/system-prompts)
- [XML-Tags zur Strukturierung Ihrer Prompts verwenden](https://docs.anthropic.com/de/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Texterstellung](https://docs.anthropic.com/de/docs/build-with-claude/text-generation): Claude überzeugt bei einer Vielzahl von textbasierten Aufgaben. Claude wurde darauf trainiert, Code, Prosa und andere natürlichsprachliche Eingaben zu verarbeiten und darauf mit Textausgaben zu reagieren.
- [Token-Zählung (Beta)](https://docs.anthropic.com/de/docs/build-with-claude/token-counting)
- [Werkzeugnutzung (Funktionsaufrufe)](https://docs.anthropic.com/de/docs/build-with-claude/tool-use)
- [Vision](https://docs.anthropic.com/de/docs/build-with-claude/vision): Die Claude 3 Modellfamilie kommt mit neuen Vision-Fähigkeiten, die es Claude ermöglichen, Bilder zu verstehen und zu analysieren, was spannende Möglichkeiten für multimodale Interaktion eröffnet.
- [Erste Einrichtung](https://docs.anthropic.com/de/docs/initial-setup): Lernen wir, wie man die Anthropic API nutzt, um mit Claude zu entwickeln.
- [Einführung in Claude](https://docs.anthropic.com/de/docs/intro-to-claude): Claude ist eine Familie von [hochleistungsfähigen und intelligenten KI-Modellen](/de/docs/about-claude/models), die von Anthropic entwickelt wurde. Während Claude leistungsstark und erweiterbar ist, ist es auch die vertrauenswürdigste und zuverlässigste verfügbare KI. Sie folgt kritischen Protokollen, macht weniger Fehler und ist resistent gegen Jailbreaks - was [Unternehmenskunden](https://www.anthropic.com/customers) ermöglicht, die sichersten KI-gestützten Anwendungen im großen Maßstab zu entwickeln.
- [Anthropic Datenschutzerklärung](https://docs.anthropic.com/de/docs/legal-center/privacy)
- [Anthropic Kochbuch](https://docs.anthropic.com/de/docs/resources/cookbook)
- [Anthropic Kurse](https://docs.anthropic.com/de/docs/resources/courses)
- [Glossar](https://docs.anthropic.com/de/docs/resources/glossary): Diese Konzepte sind nicht einzigartig für die Sprachmodelle von Anthropic, aber wir präsentieren hier eine kurze Zusammenfassung wichtiger Begriffe.
- [Claude 3 Modellkarte](https://docs.anthropic.com/de/docs/resources/model-card)
- [Modell-Abkündigungen](https://docs.anthropic.com/de/docs/resources/model-deprecations)
- [Systemstatus](https://docs.anthropic.com/de/docs/resources/status)
- [Verwendung des Evaluierungstools](https://docs.anthropic.com/de/docs/test-and-evaluate/eval-tool): Die [Anthropic Console](https://console.anthropic.com/dashboard) bietet ein **Evaluierungstool**, mit dem Sie Ihre Prompts unter verschiedenen Szenarien testen können.
- [Ausgabekonsistenz erhöhen (JSON-Modus)](https://docs.anthropic.com/de/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Claude mit Rollenprompting und Vorbefüllung im Charakter halten](https://docs.anthropic.com/de/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Jailbreaks und Prompt-Injektionen verhindern](https://docs.anthropic.com/de/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Halluzinationen reduzieren](https://docs.anthropic.com/de/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Latenz reduzieren](https://docs.anthropic.com/de/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Prompt-Leaks reduzieren](https://docs.anthropic.com/de/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Willkommen bei Claude](https://docs.anthropic.com/de/docs/welcome): Claude ist eine hochleistungsfähige, vertrauenswürdige und intelligente KI-Plattform von Anthropic. Claude überzeugt bei Aufgaben in den Bereichen Sprache, logisches Denken, Analyse, Programmierung und mehr.
- [null](https://docs.anthropic.com/de/home)
- [Adaptiver Editor](https://docs.anthropic.com/de/prompt-library/adaptive-editor): Text gemäß Benutzeranweisungen umschreiben, zum Beispiel mit anderem Tonfall, Zielgruppe oder Stil.
- [Flughafen-Code-Analyst](https://docs.anthropic.com/de/prompt-library/airport-code-analyst): Finden und extrahieren Sie Flughafen-Codes aus Text.
- [Außerirdischer Anthropologe](https://docs.anthropic.com/de/prompt-library/alien-anthropologist): Analysieren Sie die menschliche Kultur und Bräuche aus der Perspektive eines außerirdischen Anthropologen.
- [Alliterations-Alchemist](https://docs.anthropic.com/de/prompt-library/alliteration-alchemist): Generiere alliterierende Phrasen und Sätze für jedes beliebige Thema.
- [Babels Übertragungen](https://docs.anthropic.com/de/prompt-library/babels-broadcasts): Erstellen Sie überzeugende Produktankündigungen als Tweets in den 10 meistgesprochenen Sprachen der Welt.
- [Markenentwickler](https://docs.anthropic.com/de/prompt-library/brand-builder): Erstellen Sie ein Designkonzept für eine ganzheitliche Markenidentität.
- [Karrierecoach](https://docs.anthropic.com/de/prompt-library/career-coach): Führen Sie Rollenspiel-Gespräche mit einem KI-Karrierecoach.
- [Zitieren Sie Ihre Quellen](https://docs.anthropic.com/de/prompt-library/cite-your-sources): Erhalten Sie Antworten auf Fragen zum Inhalt eines Dokuments mit relevanten Zitaten zur Unterstützung der Antwort.
- [Code-Erklärer](https://docs.anthropic.com/de/prompt-library/code-clarifier): Vereinfachen und erklären Sie komplexen Code in einfacher Sprache.
- [Code-Berater](https://docs.anthropic.com/de/prompt-library/code-consultant): Schlägt Verbesserungen zur Optimierung der Python-Code-Performance vor.
- [Unternehmensweissager](https://docs.anthropic.com/de/prompt-library/corporate-clairvoyant): Extrahieren Sie Erkenntnisse, identifizieren Sie Risiken und destillieren Sie wichtige Informationen aus langen Unternehmensberichten in ein einziges Memo.
- [Kosmische Tastenanschläge](https://docs.anthropic.com/de/prompt-library/cosmic-keystrokes): Erstellen Sie ein interaktives Schnellschreibspiel in einer einzigen HTML-Datei mit seitlich scrollendem Gameplay und Tailwind CSS-Styling.
- [CSV-Konverter](https://docs.anthropic.com/de/prompt-library/csv-converter): Konvertieren Sie Daten aus verschiedenen Formaten (JSON, XML, etc.) in korrekt formatierte CSV-Dateien.
- [Kulinarischer Kreator](https://docs.anthropic.com/de/prompt-library/culinary-creator): Schlägt Rezeptideen basierend auf den verfügbaren Zutaten und Ernährungsvorlieben des Benutzers vor.
- [Datenorganisator](https://docs.anthropic.com/de/prompt-library/data-organizer): Verwandeln Sie unstrukturierten Text in maßgeschneiderte JSON-Tabellen.
- [Richtungsdekodierer](https://docs.anthropic.com/de/prompt-library/direction-decoder): Verwandeln Sie natürliche Sprache in schrittweise Anweisungen.
- [Traumdeuter](https://docs.anthropic.com/de/prompt-library/dream-interpreter): Bietet Interpretationen und Einblicke in die Symbolik der Träume des Benutzers.
- [Effizienzschätzer](https://docs.anthropic.com/de/prompt-library/efficiency-estimator): Berechnen Sie die Zeitkomplexität von Funktionen und Algorithmen.
- [E-Mail-Extraktor](https://docs.anthropic.com/de/prompt-library/email-extractor): Extrahiert E-Mail-Adressen aus einem Dokument in eine JSON-formatierte Liste.
- [Emoji-Kodierer](https://docs.anthropic.com/de/prompt-library/emoji-encoder): Wandeln Sie einfachen Text in lustige und ausdrucksstarke Emoji-Nachrichten um.
- [Ethischer Dilemma-Navigator](https://docs.anthropic.com/de/prompt-library/ethical-dilemma-navigator): Hilft dem Benutzer, komplexe ethische Dilemmata zu durchdenken und verschiedene Perspektiven zu bieten.
- [Excel-Formel-Experte](https://docs.anthropic.com/de/prompt-library/excel-formula-expert): Erstellen Sie Excel-Formeln basierend auf benutzerdefinierten Berechnungen oder Datenmanipulationen.
- [Funktionsgenerator](https://docs.anthropic.com/de/prompt-library/function-fabricator): Erstellen Sie Python-Funktionen basierend auf detaillierten Spezifikationen.
- [Futuristischer Mode-Berater](https://docs.anthropic.com/de/prompt-library/futuristic-fashion-advisor): Schlägt avantgardistische Modetrends und Stile basierend auf den spezifischen Vorlieben des Nutzers vor.
- [Git beherrschen](https://docs.anthropic.com/de/prompt-library/git-gud): Generiere passende Git-Befehle basierend auf benutzerdefinierten Versionskontrollaktionen.
- [Google Apps Scripter](https://docs.anthropic.com/de/prompt-library/google-apps-scripter): Generiere Google Apps Skripte zur Erfüllung von Aufgaben basierend auf Benutzeranforderungen.
- [Bewertungs-Guru](https://docs.anthropic.com/de/prompt-library/grading-guru): Vergleichen und bewerten Sie die Qualität von geschriebenen Texten basierend auf benutzerdefinierten Kriterien und Standards.
- [Grammatik-Genie](https://docs.anthropic.com/de/prompt-library/grammar-genie): Verwandelt grammatikalisch falsche Sätze in korrektes Englisch.
- [Hal der humorvolle Helfer](https://docs.anthropic.com/de/prompt-library/hal-the-humorous-helper): Chatte mit einer sachkundigen KI, die eine sarkastische Seite hat.
- [Redewendungs-Erklärer](https://docs.anthropic.com/de/prompt-library/idiom-illuminator): Erklärt die Bedeutung und den Ursprung gängiger Redewendungen und Sprichwörter.
- [Interviewfragen-Generator](https://docs.anthropic.com/de/prompt-library/interview-question-crafter): Generiere Fragen für Interviews.
- [LaTeX-Legende](https://docs.anthropic.com/de/prompt-library/latex-legend): Erstellen Sie LaTeX-Dokumente und generieren Sie Code für mathematische Gleichungen, Tabellen und mehr.
- [Unterrichtsplaner](https://docs.anthropic.com/de/prompt-library/lesson-planner): Erstellen Sie detaillierte Unterrichtspläne zu jedem Thema.
- [Bibliothek](https://docs.anthropic.com/de/prompt-library/library)
- [Moderator-Experte](https://docs.anthropic.com/de/prompt-library/master-moderator): Bewertung von Benutzereingaben auf potenziell schädliche oder illegale Inhalte.
- [Besprechungsprotokollant](https://docs.anthropic.com/de/prompt-library/meeting-scribe): Destillieren Sie Besprechungen in prägnante Zusammenfassungen einschließlich Diskussionsthemen, wichtiger Erkenntnisse und Aufgaben.
- [Memo-Meister](https://docs.anthropic.com/de/prompt-library/memo-maestro): Erstellen Sie umfassende Firmenmemos basierend auf Schlüsselpunkten.
- [Achtsamkeits-Mentor](https://docs.anthropic.com/de/prompt-library/mindfulness-mentor): Führen Sie den Benutzer durch Achtsamkeitsübungen und Techniken zur Stressreduktion.
- [Stimmungs-Kolorierung](https://docs.anthropic.com/de/prompt-library/mood-colorizer): Verwandelt Textbeschreibungen von Stimmungen in entsprechende HEX-Codes.
- [Motivierende Muse](https://docs.anthropic.com/de/prompt-library/motivational-muse): Bietet personalisierte Motivationsbotschaften und Affirmationen basierend auf Benutzereingaben.
- [Neologismus-Generator](https://docs.anthropic.com/de/prompt-library/neologism-creator): Erfinde neue Wörter und liefere ihre Definitionen basierend auf benutzerdefinierten Konzepten oder Ideen.
- [Perspektiven-Abwäger](https://docs.anthropic.com/de/prompt-library/perspectives-ponderer): Wägen Sie die Vor- und Nachteile eines vom Benutzer vorgegebenen Themas ab.
- [Philosophische Überlegungen](https://docs.anthropic.com/de/prompt-library/philosophical-musings): Führen Sie tiefgründige philosophische Diskussionen und Gedankenexperimente.
- [PII-Bereiniger](https://docs.anthropic.com/de/prompt-library/pii-purifier): Automatische Erkennung und Entfernung von personenbezogenen Daten (PII) aus Textdokumenten.
- [Polyglot-Superkräfte](https://docs.anthropic.com/de/prompt-library/polyglot-superpowers): Text aus jeder Sprache in jede andere Sprache übersetzen.
- [Kofferwort-Poet](https://docs.anthropic.com/de/prompt-library/portmanteau-poet): Verschmelze zwei Wörter zu einem neuen, sinnvollen Kofferwort.
- [Produkt-Namensprofi](https://docs.anthropic.com/de/prompt-library/product-naming-pro): Erstellen Sie einprägsame Produktnamen aus Beschreibungen und Schlüsselwörtern.
- [Textoptimierer](https://docs.anthropic.com/de/prompt-library/prose-polisher): Verfeinern und verbessern Sie geschriebene Inhalte mit fortgeschrittenen Lektoratstechniken und Vorschlägen.
- [Wortspiel-Experte](https://docs.anthropic.com/de/prompt-library/pun-dit): Generiere clevere Wortspiele und Wortwitze zu jedem beliebigen Thema.
- [Python Fehlerbeheber](https://docs.anthropic.com/de/prompt-library/python-bug-buster): Erkennen und beheben Sie Fehler in Python-Code.
- [Bewertungsklassifizierer](https://docs.anthropic.com/de/prompt-library/review-classifier): Kategorisiere Feedback in vordefinierte Tags und Klassifizierungen.
- [Rate mal](https://docs.anthropic.com/de/prompt-library/riddle-me-this): Generiere Rätsel und führe den Benutzer zu den Lösungen.
- [Science-Fiction-Szenario-Simulator](https://docs.anthropic.com/de/prompt-library/sci-fi-scenario-simulator): Diskutieren Sie mit dem Benutzer verschiedene Science-Fiction-Szenarien und die damit verbundenen Herausforderungen und Überlegungen.
- [Vereinfacher für die zweite Klasse](https://docs.anthropic.com/de/prompt-library/second-grade-simplifier): Mache komplexe Texte für junge Lernende leicht verständlich
- [Vergleichs-Experte](https://docs.anthropic.com/de/prompt-library/simile-savant): Generiere Vergleiche aus einfachen Beschreibungen.
- [Sokratischer Weiser](https://docs.anthropic.com/de/prompt-library/socratic-sage): Führen Sie ein sokratisches Gespräch über ein vom Benutzer vorgegebenes Thema.
- [Tabellenkalkulationszauberer](https://docs.anthropic.com/de/prompt-library/spreadsheet-sorcerer): Generiere CSV-Tabellenkalkulationen mit verschiedenen Datentypen.
- [SQL-Zauberer](https://docs.anthropic.com/de/prompt-library/sql-sorcerer): Verwandelt Alltagssprache in SQL-Abfragen.
- [Geschichten-Erzählpartner](https://docs.anthropic.com/de/prompt-library/storytelling-sidekick): Erstellen Sie gemeinsam mit dem Benutzer fesselnde Geschichten und bieten Sie Handlungswendungen und Charakterentwicklung an.
- [Zeitreise-Berater](https://docs.anthropic.com/de/prompt-library/time-travel-consultant): Hilft dem Benutzer bei der Navigation durch hypothetische Zeitreise-Szenarien und deren Auswirkungen.
- [Zungenbrecher](https://docs.anthropic.com/de/prompt-library/tongue-twister): Erstellen Sie herausfordernde Zungenbrecher.
- [Quiz-Generator](https://docs.anthropic.com/de/prompt-library/trivia-generator): Generiere Quizfragen zu verschiedenen Themen und biete bei Bedarf Hinweise an.
- [Tweet-Tonalitätsdetektor](https://docs.anthropic.com/de/prompt-library/tweet-tone-detector): Erkennen Sie den Ton und die Stimmung hinter Tweets.
- [VR-Fitness-Innovator](https://docs.anthropic.com/de/prompt-library/vr-fitness-innovator): Kreative Ideen für Virtual-Reality-Fitness-Spiele entwickeln.
- [Website-Assistent](https://docs.anthropic.com/de/prompt-library/website-wizard): Erstellen Sie einseitige Websites basierend auf Benutzervorgaben.
- [API](https://docs.anthropic.com/de/release-notes/api): Verfolgen Sie Updates der Anthropic API und der Entwicklerkonsole.
- [Claude Apps](https://docs.anthropic.com/de/release-notes/claude-apps): Verfolgen Sie Updates über Anthropics Claude-Anwendungen.
- [Überblick](https://docs.anthropic.com/de/release-notes/overview): Bleiben Sie auf dem Laufenden über Updates für Anthropics Produkte und Dienstleistungen.
- [Systemaufforderungen](https://docs.anthropic.com/de/release-notes/system-prompts): Sehen Sie Aktualisierungen der Kern-Systemaufforderungen auf [Claude.ai](https://www.claude.ai) und den Claude [iOS](http://anthropic.com/ios) und [Android](http://anthropic.com/android) Apps.
- [Cancel a Message Batch (beta)](https://docs.anthropic.com/en/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/en/api/claude-on-amazon-bedrock): Anthropic’s Claude models are now generally available through Amazon Bedrock.
- [Vertex AI API](https://docs.anthropic.com/en/api/claude-on-vertex-ai): Anthropic’s Claude models are now generally available through [Vertex AI](https://cloud.google.com/vertex-ai).
- [Client SDKs](https://docs.anthropic.com/en/api/client-sdks): We provide libraries in Python and TypeScript that make it easier to work with the Anthropic API.
- [Create a Text Completion](https://docs.anthropic.com/en/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Create a Message Batch (beta)](https://docs.anthropic.com/en/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Errors](https://docs.anthropic.com/en/api/errors)
- [Getting help](https://docs.anthropic.com/en/api/getting-help): We've tried to provide the answers to the most common questions in these docs. However, if you need further technical support using Claude, the Anthropic API, or any of our products, you may reach our support team at [support.anthropic.com](https://support.anthropic.com).
- [Getting started](https://docs.anthropic.com/en/api/getting-started)
- [IP addresses](https://docs.anthropic.com/en/api/ip-addresses): Anthropic services live at a fixed range of IP addresses. You can add these to your firewall to open the minimum amount of surface area for egress traffic when accessing the Anthropic API and Console. These ranges will not change without notice.
- [List Message Batches (beta)](https://docs.anthropic.com/en/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/en/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Message Batches examples](https://docs.anthropic.com/en/api/messages-batch-examples): Example usage for the Message Batches API
- [Count Message tokens (beta)](https://docs.anthropic.com/en/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Messages examples](https://docs.anthropic.com/en/api/messages-examples): Request and response examples for the Messages API
- [Streaming Messages](https://docs.anthropic.com/en/api/messages-streaming)
- [Migrating from Text Completions](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages): Migrating from Text Completions to Messages
- [Prompt validation](https://docs.anthropic.com/en/api/prompt-validation): With Text Completions
- [Rate limits](https://docs.anthropic.com/en/api/rate-limits): To mitigate against misuse and manage capacity on our API, we have implemented limits on how much an organization can use the Claude API.
- [Retrieve Message Batch Results (beta)](https://docs.anthropic.com/en/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Retrieve a Message Batch (beta)](https://docs.anthropic.com/en/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Streaming Text Completions](https://docs.anthropic.com/en/api/streaming)
- [Supported regions](https://docs.anthropic.com/en/api/supported-regions): Here are the countries, regions, and territories we can currently support access from:
- [Versions](https://docs.anthropic.com/en/api/versioning): When making API requests, you must send an `anthropic-version` request header. For example, `anthropic-version: 2023-06-01`. If you are using our [client libraries](/en/api/client-libraries), this is handled for you automatically.
- [July 2024 Updates](https://docs.anthropic.com/en/developer-newsletter/july2024): July 28, 2024
- [Overview](https://docs.anthropic.com/en/developer-newsletter/overview): Explore monthly updates, engineering deep dives, best practices, and success stories to enhance your Claude integrations
- [Models](https://docs.anthropic.com/en/docs/about-claude/models): Claude is a family of state-of-the-art large language models developed by Anthropic. This guide introduces our models and compares their performance with legacy models. 
- [Security and compliance](https://docs.anthropic.com/en/docs/about-claude/security-compliance)
- [Content moderation](https://docs.anthropic.com/en/docs/about-claude/use-case-guides/content-moderation): Content moderation is a critical aspect of maintaining a safe, respectful, and productive environment in digital applications. In this guide, we'll discuss how Claude can be used to moderate content within your digital application.
- [Customer support agent](https://docs.anthropic.com/en/docs/about-claude/use-case-guides/customer-support-chat): This guide walks through how to leverage Claude's advanced conversational capabilities to handle customer inquiries in real time, providing 24/7 support, reducing wait times, and managing high support volumes with accurate responses and positive interactions.
- [Legal summarization](https://docs.anthropic.com/en/docs/about-claude/use-case-guides/legal-summarization): This guide walks through how to leverage Claude's advanced natural language processing capabilities to efficiently summarize legal documents, extracting key information and expediting legal research. With Claude, you can streamline the review of contracts, litigation prep, and regulatory work, saving time and ensuring accuracy in your legal processes.
- [Guides to common use cases](https://docs.anthropic.com/en/docs/about-claude/use-case-guides/overview)
- [Ticket routing](https://docs.anthropic.com/en/docs/about-claude/use-case-guides/ticket-routing): This guide walks through how to harness Claude's advanced natural language understanding capabilities to classify customer support tickets at scale based on customer intent, urgency, prioritization, customer profile, and more.
- [Google Sheets add-on](https://docs.anthropic.com/en/docs/build-with-claude/claude-for-sheets): The [Claude for Sheets extension](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) integrates Claude into Google Sheets, allowing you to execute interactions with Claude directly in cells.
- [Computer use (beta)](https://docs.anthropic.com/en/docs/build-with-claude/computer-use)
- [Define your success criteria](https://docs.anthropic.com/en/docs/build-with-claude/define-success)
- [Create strong empirical evaluations](https://docs.anthropic.com/en/docs/build-with-claude/develop-tests)
- [Embeddings](https://docs.anthropic.com/en/docs/build-with-claude/embeddings): Text embeddings are numerical representations of text that enable measuring semantic similarity. This guide introduces embeddings, their applications, and how to use embedding models for tasks like search, recommendations, and anomaly detection.
- [Message Batches (beta)](https://docs.anthropic.com/en/docs/build-with-claude/message-batches)
- [PDF support (beta)](https://docs.anthropic.com/en/docs/build-with-claude/pdf-support): The Claude 3.5 Sonnet models now support PDF input and understand both text and visual content within documents.
- [Prompt caching (beta)](https://docs.anthropic.com/en/docs/build-with-claude/prompt-caching)
- [Be clear, direct, and detailed](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Let Claude think (chain of thought prompting) to increase performance](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Chain complex prompts for stronger performance](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Long context prompting tips](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Use examples (multishot prompting) to guide Claude's behavior](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Prompt engineering overview](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/overview)
- [Prefill Claude's response for greater output control](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Automatically generate first draft prompt templates](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Use our prompt improver to optimize your prompts](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Use prompt templates and variables](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Giving Claude a role with a system prompt](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/system-prompts)
- [Use XML tags to structure your prompts](https://docs.anthropic.com/en/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Text generation](https://docs.anthropic.com/en/docs/build-with-claude/text-generation): Claude excels in a wide range of text-based tasks. Claude has been trained to ingest code, prose, and other natural language inputs, and provide text outputs in response.
- [Token counting (beta)](https://docs.anthropic.com/en/docs/build-with-claude/token-counting)
- [Tool use (function calling)](https://docs.anthropic.com/en/docs/build-with-claude/tool-use)
- [Vision](https://docs.anthropic.com/en/docs/build-with-claude/vision): The Claude 3 family of models comes with new vision capabilities that allow Claude to understand and analyze images, opening up exciting possibilities for multimodal interaction.
- [Initial setup](https://docs.anthropic.com/en/docs/initial-setup): Let’s learn how to use the Anthropic API to build with Claude.
- [Intro to Claude](https://docs.anthropic.com/en/docs/intro-to-claude): Claude is a family of [highly performant and intelligent AI models](/en/docs/about-claude/models) built by Anthropic. While Claude is powerful and extensible, it’s also the most trustworthy and reliable AI available. It follows critical protocols, makes fewer mistakes, and is resistant to jailbreaks—allowing [enterprise customers](https://www.anthropic.com/customers) to build the safest AI-powered applications at scale.
- [Anthropic Privacy Policy](https://docs.anthropic.com/en/docs/legal-center/privacy)
- [Anthropic Cookbook](https://docs.anthropic.com/en/docs/resources/cookbook)
- [Anthropic Courses](https://docs.anthropic.com/en/docs/resources/courses)
- [Glossary](https://docs.anthropic.com/en/docs/resources/glossary): These concepts are not unique to Anthropic’s language models, but we present a brief summary of key terms below.
- [Claude 3 model card](https://docs.anthropic.com/en/docs/resources/model-card)
- [Model Deprecations](https://docs.anthropic.com/en/docs/resources/model-deprecations)
- [System status](https://docs.anthropic.com/en/docs/resources/status)
- [Using the Evaluation Tool](https://docs.anthropic.com/en/docs/test-and-evaluate/eval-tool): The [Anthropic Console](https://console.anthropic.com/dashboard) features an **Evaluation tool** that allows you to test your prompts under various scenarios.
- [Increase output consistency (JSON mode)](https://docs.anthropic.com/en/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Keep Claude in character with role prompting and prefilling](https://docs.anthropic.com/en/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Mitigate jailbreaks and prompt injections](https://docs.anthropic.com/en/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Reduce hallucinations](https://docs.anthropic.com/en/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Reducing latency](https://docs.anthropic.com/en/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Reduce prompt leak](https://docs.anthropic.com/en/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Welcome to Claude](https://docs.anthropic.com/en/docs/welcome): Claude is a highly performant, trustworthy, and intelligent AI platform built by Anthropic. Claude excels at tasks involving language, reasoning, analysis, coding, and more.
- [null](https://docs.anthropic.com/en/home)
- [Adaptive editor](https://docs.anthropic.com/en/prompt-library/adaptive-editor): Rewrite text following user-given instructions, such as with a different tone, audience, or style.
- [Airport code analyst](https://docs.anthropic.com/en/prompt-library/airport-code-analyst): Find and extract airport codes from text.
- [Alien anthropologist](https://docs.anthropic.com/en/prompt-library/alien-anthropologist): Analyze human culture and customs from the perspective of an alien anthropologist.
- [Alliteration alchemist](https://docs.anthropic.com/en/prompt-library/alliteration-alchemist): Generate alliterative phrases and sentences for any given subject.
- [Babel's broadcasts](https://docs.anthropic.com/en/prompt-library/babels-broadcasts): Create compelling product announcement tweets in the world's 10 most spoken languages.
- [Brand builder](https://docs.anthropic.com/en/prompt-library/brand-builder): Craft a design brief for a holistic brand identity.
- [Career coach](https://docs.anthropic.com/en/prompt-library/career-coach): Engage in role-play conversations with an AI career coach.
- [Cite your sources](https://docs.anthropic.com/en/prompt-library/cite-your-sources): Get answers to questions about a document's content with relevant citations supporting the response.
- [Code clarifier](https://docs.anthropic.com/en/prompt-library/code-clarifier): Simplify and explain complex code in plain language.
- [Code consultant](https://docs.anthropic.com/en/prompt-library/code-consultant): Suggest improvements to optimize Python code performance.
- [Corporate clairvoyant](https://docs.anthropic.com/en/prompt-library/corporate-clairvoyant): Extract insights, identify risks, and distill key information from long corporate reports into a single memo.
- [Cosmic Keystrokes](https://docs.anthropic.com/en/prompt-library/cosmic-keystrokes): Generate an interactive speed typing game in a single HTML file, featuring side-scrolling gameplay and Tailwind CSS styling.
- [CSV converter](https://docs.anthropic.com/en/prompt-library/csv-converter): Convert data from various formats (JSON, XML, etc.) into properly formatted CSV files.
- [Culinary creator](https://docs.anthropic.com/en/prompt-library/culinary-creator): Suggest recipe ideas based on the user's available ingredients and dietary preferences.
- [Data organizer](https://docs.anthropic.com/en/prompt-library/data-organizer): Turn unstructured text into bespoke JSON tables.
- [Direction decoder](https://docs.anthropic.com/en/prompt-library/direction-decoder): Transform natural language into step-by-step directions.
- [Dream interpreter](https://docs.anthropic.com/en/prompt-library/dream-interpreter): Offer interpretations and insights into the symbolism of the user's dreams.
- [Efficiency estimator](https://docs.anthropic.com/en/prompt-library/efficiency-estimator): Calculate the time complexity of functions and algorithms.
- [Email extractor](https://docs.anthropic.com/en/prompt-library/email-extractor): Extract email addresses from a document into a JSON-formatted list.
- [Emoji encoder](https://docs.anthropic.com/en/prompt-library/emoji-encoder): Convert plain text into fun and expressive emoji messages.
- [Ethical dilemma navigator](https://docs.anthropic.com/en/prompt-library/ethical-dilemma-navigator): Help the user think through complex ethical dilemmas and provide different perspectives.
- [Excel formula expert](https://docs.anthropic.com/en/prompt-library/excel-formula-expert): Create Excel formulas based on user-described calculations or data manipulations.
- [Function fabricator](https://docs.anthropic.com/en/prompt-library/function-fabricator): Create Python functions based on detailed specifications.
- [Futuristic fashion advisor](https://docs.anthropic.com/en/prompt-library/futuristic-fashion-advisor): Suggest avant-garde fashion trends and styles for the user's specific preferences.
- [Git gud](https://docs.anthropic.com/en/prompt-library/git-gud): Generate appropriate Git commands based on user-described version control actions.
- [Google apps scripter](https://docs.anthropic.com/en/prompt-library/google-apps-scripter): Generate Google Apps scripts to complete tasks based on user requirements.
- [Grading guru](https://docs.anthropic.com/en/prompt-library/grading-guru): Compare and evaluate the quality of written texts based on user-defined criteria and standards.
- [Grammar genie](https://docs.anthropic.com/en/prompt-library/grammar-genie): Transform grammatically incorrect sentences into proper English.
- [Hal the humorous helper](https://docs.anthropic.com/en/prompt-library/hal-the-humorous-helper): Chat with a knowledgeable AI that has a sarcastic side.
- [Idiom illuminator](https://docs.anthropic.com/en/prompt-library/idiom-illuminator): Explain the meaning and origin of common idioms and proverbs.
- [Interview question crafter](https://docs.anthropic.com/en/prompt-library/interview-question-crafter): Generate questions for interviews.
- [LaTeX legend](https://docs.anthropic.com/en/prompt-library/latex-legend): Write LaTeX documents, generating code for mathematical equations, tables, and more.
- [Lesson planner](https://docs.anthropic.com/en/prompt-library/lesson-planner): Craft in depth lesson plans on any subject.
- [Library](https://docs.anthropic.com/en/prompt-library/library)
- [Master moderator](https://docs.anthropic.com/en/prompt-library/master-moderator): Evaluate user inputs for potential harmful or illegal content.
- [Meeting scribe](https://docs.anthropic.com/en/prompt-library/meeting-scribe): Distill meetings into concise summaries including discussion topics, key takeaways, and action items.
- [Memo maestro](https://docs.anthropic.com/en/prompt-library/memo-maestro): Compose comprehensive company memos based on key points.
- [Mindfulness mentor](https://docs.anthropic.com/en/prompt-library/mindfulness-mentor): Guide the user through mindfulness exercises and techniques for stress reduction.
- [Mood colorizer](https://docs.anthropic.com/en/prompt-library/mood-colorizer): Transform text descriptions of moods into corresponding HEX codes.
- [Motivational muse](https://docs.anthropic.com/en/prompt-library/motivational-muse): Provide personalized motivational messages and affirmations based on user input.
- [Neologism creator](https://docs.anthropic.com/en/prompt-library/neologism-creator): Invent new words and provide their definitions based on user-provided concepts or ideas.
- [Perspectives ponderer](https://docs.anthropic.com/en/prompt-library/perspectives-ponderer): Weigh the pros and cons of a user-provided topic.
- [Philosophical musings](https://docs.anthropic.com/en/prompt-library/philosophical-musings): Engage in deep philosophical discussions and thought experiments.
- [PII purifier](https://docs.anthropic.com/en/prompt-library/pii-purifier): Automatically detect and remove personally identifiable information (PII) from text documents.
- [Polyglot superpowers](https://docs.anthropic.com/en/prompt-library/polyglot-superpowers): Translate text from any language into any language.
- [Portmanteau poet](https://docs.anthropic.com/en/prompt-library/portmanteau-poet): Blend two words together to create a new, meaningful portmanteau.
- [Product naming pro](https://docs.anthropic.com/en/prompt-library/product-naming-pro): Create catchy product names from descriptions and keywords.
- [Prose polisher](https://docs.anthropic.com/en/prompt-library/prose-polisher): Refine and improve written content with advanced copyediting techniques and suggestions.
- [Pun-dit](https://docs.anthropic.com/en/prompt-library/pun-dit): Generate clever puns and wordplay based on any given topic.
- [Python bug buster](https://docs.anthropic.com/en/prompt-library/python-bug-buster): Detect and fix bugs in Python code.
- [Review classifier](https://docs.anthropic.com/en/prompt-library/review-classifier): Categorize feedback into pre-specified tags and categorizations.
- [Riddle me this](https://docs.anthropic.com/en/prompt-library/riddle-me-this): Generate riddles and guide the user to the solutions.
- [Sci-fi scenario simulator](https://docs.anthropic.com/en/prompt-library/sci-fi-scenario-simulator): Discuss with the user various science fiction scenarios and associated challenges and considerations.
- [Second-grade simplifier](https://docs.anthropic.com/en/prompt-library/second-grade-simplifier): Make complex text easy for young learners to understand.
- [Simile savant](https://docs.anthropic.com/en/prompt-library/simile-savant): Generate similes from basic descriptions.
- [Socratic sage](https://docs.anthropic.com/en/prompt-library/socratic-sage): Engage in Socratic style conversation over a user-given topic.
- [Spreadsheet sorcerer](https://docs.anthropic.com/en/prompt-library/spreadsheet-sorcerer): Generate CSV spreadsheets with various types of data.
- [SQL sorcerer](https://docs.anthropic.com/en/prompt-library/sql-sorcerer): Transform everyday language into SQL queries.
- [Storytelling sidekick](https://docs.anthropic.com/en/prompt-library/storytelling-sidekick): Collaboratively create engaging stories with the user, offering plot twists and character development.
- [Time travel consultant](https://docs.anthropic.com/en/prompt-library/time-travel-consultant): Help the user navigate hypothetical time travel scenarios and their implications.
- [Tongue twister](https://docs.anthropic.com/en/prompt-library/tongue-twister): Create challenging tongue twisters.
- [Trivia generator](https://docs.anthropic.com/en/prompt-library/trivia-generator): Generate trivia questions on a wide range of topics and provide hints when needed.
- [Tweet tone detector](https://docs.anthropic.com/en/prompt-library/tweet-tone-detector): Detect the tone and sentiment behind tweets.
- [VR fitness innovator](https://docs.anthropic.com/en/prompt-library/vr-fitness-innovator): Brainstorm creative ideas for virtual reality fitness games.
- [Website wizard](https://docs.anthropic.com/en/prompt-library/website-wizard): Create one-page websites based on user specifications.
- [API](https://docs.anthropic.com/en/release-notes/api): Follow along with updates across Anthropic's API and Developer Console.
- [Claude Apps](https://docs.anthropic.com/en/release-notes/claude-apps): Follow along with updates across Anthropic's Claude applications.
- [Overview](https://docs.anthropic.com/en/release-notes/overview): Follow along with updates across Anthropic's products and services.
- [System Prompts](https://docs.anthropic.com/en/release-notes/system-prompts): See updates to the core system prompts on [Claude.ai](https://www.claude.ai) and the Claude [iOS](http://anthropic.com/ios) and [Android](http://anthropic.com/android) apps.
- [Cancelar un Lote de Mensajes (beta)](https://docs.anthropic.com/es/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [API de Amazon Bedrock](https://docs.anthropic.com/es/api/claude-on-amazon-bedrock): Los modelos Claude de Anthropic están ahora generalmente disponibles a través de Amazon Bedrock.
- [API de Vertex AI](https://docs.anthropic.com/es/api/claude-on-vertex-ai): Los modelos Claude de Anthropic ahora están generalmente disponibles a través de [Vertex AI](https://cloud.google.com/vertex-ai).
- [SDKs de Cliente](https://docs.anthropic.com/es/api/client-sdks): Proporcionamos bibliotecas en Python y TypeScript que facilitan el trabajo con la API de Anthropic.
- [Create a Text Completion](https://docs.anthropic.com/es/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Crear un Lote de Mensajes (beta)](https://docs.anthropic.com/es/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Errores](https://docs.anthropic.com/es/api/errors)
- [Obtener ayuda](https://docs.anthropic.com/es/api/getting-help): Hemos intentado proporcionar las respuestas a las preguntas más comunes en esta documentación. Sin embargo, si necesita más soporte técnico usando Claude, la API de Anthropic, o cualquiera de nuestros productos, puede contactar a nuestro equipo de soporte en [support.anthropic.com](https://support.anthropic.com).
- [Primeros pasos](https://docs.anthropic.com/es/api/getting-started)
- [Direcciones IP](https://docs.anthropic.com/es/api/ip-addresses): Los servicios de Anthropic se encuentran en un rango fijo de direcciones IP. Puede agregar estas a su firewall para abrir la mínima superficie de área para el tráfico de salida al acceder a la API y Consola de Anthropic. Estos rangos no cambiarán sin previo aviso.
- [Listar Lotes de Mensajes (beta)](https://docs.anthropic.com/es/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/es/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Ejemplos de Lotes de Mensajes](https://docs.anthropic.com/es/api/messages-batch-examples): Ejemplos de uso para la API de Lotes de Mensajes
- [Contar tokens de Mensaje (beta)](https://docs.anthropic.com/es/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Ejemplos de mensajes](https://docs.anthropic.com/es/api/messages-examples): Ejemplos de solicitudes y respuestas para la API de Mensajes
- [Mensajes en Streaming](https://docs.anthropic.com/es/api/messages-streaming)
- [Migrando desde Completaciones de Texto](https://docs.anthropic.com/es/api/migrating-from-text-completions-to-messages): Migrando desde Completaciones de Texto a Mensajes
- [Validación de prompts](https://docs.anthropic.com/es/api/prompt-validation): Con Text Completions
- [Límites de uso](https://docs.anthropic.com/es/api/rate-limits): Para mitigar el mal uso y gestionar la capacidad en nuestra API, hemos implementado límites sobre cuánto puede usar una organización la API de Claude.
- [Obtener Resultados de Lote de Mensajes (beta)](https://docs.anthropic.com/es/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Recuperar un Lote de Mensajes (beta)](https://docs.anthropic.com/es/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Transmisión de Completados de Texto](https://docs.anthropic.com/es/api/streaming)
- [Regiones admitidas](https://docs.anthropic.com/es/api/supported-regions): Aquí están los países, regiones y territorios desde los que actualmente podemos admitir el acceso:
- [Versiones](https://docs.anthropic.com/es/api/versioning): Al realizar solicitudes a la API, debes enviar un encabezado de solicitud `anthropic-version`. Por ejemplo, `anthropic-version: 2023-06-01`. Si estás utilizando nuestras [bibliotecas cliente](/es/api/client-libraries), esto se maneja automáticamente por ti.
- [Actualizaciones de Julio 2024](https://docs.anthropic.com/es/developer-newsletter/july2024): 28 de julio de 2024
- [Descripción General](https://docs.anthropic.com/es/developer-newsletter/overview): Explore actualizaciones mensuales, análisis técnicos detallados, mejores prácticas e historias de éxito para mejorar sus integraciones con Claude
- [Modelos](https://docs.anthropic.com/es/docs/about-claude/models): Claude es una familia de modelos de lenguaje de última generación desarrollados por Anthropic. Esta guía presenta nuestros modelos y compara su rendimiento con modelos anteriores.
- [Seguridad y cumplimiento](https://docs.anthropic.com/es/docs/about-claude/security-compliance)
- [Moderación de contenido](https://docs.anthropic.com/es/docs/about-claude/use-case-guides/content-moderation): La moderación de contenido es un aspecto crítico para mantener un entorno seguro, respetuoso y productivo en las aplicaciones digitales. En esta guía, discutiremos cómo se puede utilizar Claude para moderar contenido dentro de su aplicación digital.
- [Agente de atención al cliente](https://docs.anthropic.com/es/docs/about-claude/use-case-guides/customer-support-chat): Esta guía explica cómo aprovechar las capacidades conversacionales avanzadas de Claude para manejar consultas de clientes en tiempo real, proporcionando soporte 24/7, reduciendo tiempos de espera y gestionando altos volúmenes de soporte con respuestas precisas e interacciones positivas.
- [Resumen legal](https://docs.anthropic.com/es/docs/about-claude/use-case-guides/legal-summarization): Esta guía explica cómo aprovechar las capacidades avanzadas de procesamiento del lenguaje natural de Claude para resumir eficientemente documentos legales, extrayendo información clave y acelerando la investigación legal. Con Claude, puede optimizar la revisión de contratos, la preparación de litigios y el trabajo regulatorio, ahorrando tiempo y garantizando la precisión en sus procesos legales.
- [Guías para casos de uso comunes](https://docs.anthropic.com/es/docs/about-claude/use-case-guides/overview)
- [Enrutamiento de tickets](https://docs.anthropic.com/es/docs/about-claude/use-case-guides/ticket-routing): Esta guía explica cómo aprovechar las capacidades avanzadas de comprensión del lenguaje natural de Claude para clasificar tickets de soporte al cliente a escala basándose en la intención del cliente, urgencia, priorización, perfil del cliente y más.
- [Complemento de Google Sheets](https://docs.anthropic.com/es/docs/build-with-claude/claude-for-sheets): La [extensión Claude for Sheets](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) integra Claude en Google Sheets, permitiéndote ejecutar interacciones con Claude directamente en las celdas.
- [null](https://docs.anthropic.com/es/docs/build-with-claude/computer-use)
- [Define sus criterios de éxito](https://docs.anthropic.com/es/docs/build-with-claude/define-success)
- [Crear evaluaciones empíricas sólidas](https://docs.anthropic.com/es/docs/build-with-claude/develop-tests)
- [Embeddings](https://docs.anthropic.com/es/docs/build-with-claude/embeddings): Los embeddings de texto son representaciones numéricas del texto que permiten medir la similitud semántica. Esta guía introduce los embeddings, sus aplicaciones y cómo usar modelos de embedding para tareas como búsqueda, recomendaciones y detección de anomalías.
- [Lotes de Mensajes (beta)](https://docs.anthropic.com/es/docs/build-with-claude/message-batches)
- [Soporte de PDF (beta)](https://docs.anthropic.com/es/docs/build-with-claude/pdf-support): Los modelos Claude 3.5 Sonnet ahora admiten entrada de PDF y comprenden tanto el contenido textual como visual dentro de los documentos.
- [Caché de prompts (beta)](https://docs.anthropic.com/es/docs/build-with-claude/prompt-caching)
- [Sé claro, directo y detallado](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Deja que Claude piense (indicaciones de cadena de pensamiento) para aumentar el rendimiento](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Encadena prompts complejos para un rendimiento más sólido](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Consejos para el uso de contexto largo](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Usar ejemplos (prompting multishot) para guiar el comportamiento de Claude](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Descripción general de la ingeniería de prompts](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/overview)
- [Prellenar la respuesta de Claude para un mayor control de la salida](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Generar automáticamente plantillas de prompts iniciales](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Usa nuestro mejorador de prompts para optimizar tus prompts](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Usar plantillas y variables de prompt](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Asignando un rol a Claude con un prompt de sistema](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/system-prompts)
- [Usa etiquetas XML para estructurar tus prompts](https://docs.anthropic.com/es/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Generación de texto](https://docs.anthropic.com/es/docs/build-with-claude/text-generation): Claude sobresale en una amplia gama de tareas basadas en texto. Claude ha sido entrenado para procesar código, prosa y otras entradas de lenguaje natural, y proporcionar salidas de texto como respuesta.
- [Conteo de tokens (beta)](https://docs.anthropic.com/es/docs/build-with-claude/token-counting)
- [Uso de herramientas (llamada a funciones)](https://docs.anthropic.com/es/docs/build-with-claude/tool-use)
- [Visión](https://docs.anthropic.com/es/docs/build-with-claude/vision): La familia de modelos Claude 3 viene con nuevas capacidades de visión que permiten a Claude entender y analizar imágenes, abriendo emocionantes posibilidades para la interacción multimodal.
- [Configuración inicial](https://docs.anthropic.com/es/docs/initial-setup): Aprendamos cómo usar la API de Anthropic para construir con Claude.
- [Introducción a Claude](https://docs.anthropic.com/es/docs/intro-to-claude): Claude es una familia de [modelos de IA altamente eficientes e inteligentes](/es/docs/about-claude/models) creados por Anthropic. Si bien Claude es potente y extensible, también es la IA más confiable y segura disponible. Sigue protocolos críticos, comete menos errores y es resistente a los intentos de evasión, permitiendo a los [clientes empresariales](https://www.anthropic.com/customers) construir las aplicaciones más seguras impulsadas por IA a escala.
- [Política de Privacidad de Anthropic](https://docs.anthropic.com/es/docs/legal-center/privacy)
- [Libro de Recetas de Anthropic](https://docs.anthropic.com/es/docs/resources/cookbook)
- [Cursos de Anthropic](https://docs.anthropic.com/es/docs/resources/courses)
- [Glosario](https://docs.anthropic.com/es/docs/resources/glossary): Estos conceptos no son exclusivos de los modelos de lenguaje de Anthropic, pero presentamos un breve resumen de términos clave a continuación.
- [Ficha técnica del modelo Claude 3](https://docs.anthropic.com/es/docs/resources/model-card)
- [Obsolescencia de Modelos](https://docs.anthropic.com/es/docs/resources/model-deprecations)
- [Estado del sistema](https://docs.anthropic.com/es/docs/resources/status)
- [Usando la Herramienta de Evaluación](https://docs.anthropic.com/es/docs/test-and-evaluate/eval-tool): La [Consola de Anthropic](https://console.anthropic.com/dashboard) incluye una **Herramienta de evaluación** que te permite probar tus prompts bajo varios escenarios.
- [Aumentar la consistencia de la salida (modo JSON)](https://docs.anthropic.com/es/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Mantener a Claude en personaje con indicaciones de rol y precompletado](https://docs.anthropic.com/es/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Mitigar los jailbreaks e inyecciones de prompt](https://docs.anthropic.com/es/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Reducir las alucinaciones](https://docs.anthropic.com/es/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Reduciendo la latencia](https://docs.anthropic.com/es/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Reducir la filtración de prompts](https://docs.anthropic.com/es/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Bienvenido a Claude](https://docs.anthropic.com/es/docs/welcome): Claude es una plataforma de IA altamente eficiente, confiable e inteligente desarrollada por Anthropic. Claude sobresale en tareas que involucran lenguaje, razonamiento, análisis, programación y más.
- [null](https://docs.anthropic.com/es/home)
- [Editor adaptativo](https://docs.anthropic.com/es/prompt-library/adaptive-editor): Reescribe texto siguiendo instrucciones del usuario, como con un tono, audiencia o estilo diferente.
- [Analizador de códigos de aeropuerto](https://docs.anthropic.com/es/prompt-library/airport-code-analyst): Encuentra y extrae códigos de aeropuerto del texto.
- [Antropólogo alienígena](https://docs.anthropic.com/es/prompt-library/alien-anthropologist): Analiza la cultura y costumbres humanas desde la perspectiva de un antropólogo alienígena.
- [Alquimista de la aliteración](https://docs.anthropic.com/es/prompt-library/alliteration-alchemist): Genera frases y oraciones con aliteración para cualquier tema dado.
- [Transmisiones de Babel](https://docs.anthropic.com/es/prompt-library/babels-broadcasts): Crea tweets convincentes de anuncios de productos en los 10 idiomas más hablados del mundo.
- [Constructor de marca](https://docs.anthropic.com/es/prompt-library/brand-builder): Crea un brief de diseño para una identidad de marca holística.
- [Entrenador profesional](https://docs.anthropic.com/es/prompt-library/career-coach): Participa en conversaciones de juego de roles con un entrenador profesional de IA.
- [Cita tus fuentes](https://docs.anthropic.com/es/prompt-library/cite-your-sources): Obtén respuestas a preguntas sobre el contenido de un documento con citas relevantes que respaldan la respuesta.
- [Clarificador de código](https://docs.anthropic.com/es/prompt-library/code-clarifier): Simplifica y explica código complejo en lenguaje sencillo.
- [Consultor de código](https://docs.anthropic.com/es/prompt-library/code-consultant): Sugerir mejoras para optimizar el rendimiento del código Python.
- [Clarividente corporativo](https://docs.anthropic.com/es/prompt-library/corporate-clairvoyant): Extrae perspectivas, identifica riesgos y destila información clave de extensos informes corporativos en un solo memorando.
- [Pulsaciones Cósmicas](https://docs.anthropic.com/es/prompt-library/cosmic-keystrokes): Genera un juego interactivo de escritura rápida en un solo archivo HTML, con jugabilidad de desplazamiento lateral y estilo Tailwind CSS.
- [Convertidor CSV](https://docs.anthropic.com/es/prompt-library/csv-converter): Convierte datos de varios formatos (JSON, XML, etc.) en archivos CSV correctamente formateados.
- [Creador culinario](https://docs.anthropic.com/es/prompt-library/culinary-creator): Sugiere ideas de recetas basadas en los ingredientes disponibles y las preferencias dietéticas del usuario.
- [Organizador de datos](https://docs.anthropic.com/es/prompt-library/data-organizer): Convierte texto no estructurado en tablas JSON personalizadas.
- [Decodificador de direcciones](https://docs.anthropic.com/es/prompt-library/direction-decoder): Transforma el lenguaje natural en instrucciones paso a paso.
- [Intérprete de sueños](https://docs.anthropic.com/es/prompt-library/dream-interpreter): Ofrece interpretaciones y perspectivas sobre el simbolismo de los sueños del usuario
- [Estimador de eficiencia](https://docs.anthropic.com/es/prompt-library/efficiency-estimator): Calcula la complejidad temporal de funciones y algoritmos.
- [Extractor de correos electrónicos](https://docs.anthropic.com/es/prompt-library/email-extractor): Extrae direcciones de correo electrónico de un documento a una lista con formato JSON.
- [Codificador de emojis](https://docs.anthropic.com/es/prompt-library/emoji-encoder): Convierte texto plano en mensajes divertidos y expresivos con emojis.
- [Navegador de dilemas éticos](https://docs.anthropic.com/es/prompt-library/ethical-dilemma-navigator): Ayuda al usuario a reflexionar sobre dilemas éticos complejos y proporciona diferentes perspectivas.
- [Experto en fórmulas de Excel](https://docs.anthropic.com/es/prompt-library/excel-formula-expert): Crea fórmulas de Excel basadas en cálculos o manipulaciones de datos descritos por el usuario.
- [Fabricante de funciones](https://docs.anthropic.com/es/prompt-library/function-fabricator): Crea funciones Python basadas en especificaciones detalladas.
- [Asesor de moda futurista](https://docs.anthropic.com/es/prompt-library/futuristic-fashion-advisor): Sugiere tendencias y estilos de moda vanguardista según las preferencias específicas del usuario
- [Domina Git](https://docs.anthropic.com/es/prompt-library/git-gud): Genera comandos Git apropiados basados en acciones de control de versiones descritas por el usuario.
- [Programador de scripts de Google Apps](https://docs.anthropic.com/es/prompt-library/google-apps-scripter): Genera scripts de Google Apps para completar tareas basadas en los requisitos del usuario.
- [Gurú de la calificación](https://docs.anthropic.com/es/prompt-library/grading-guru): Compara y evalúa la calidad de textos escritos basándose en criterios y estándares definidos por el usuario.
- [Genio gramatical](https://docs.anthropic.com/es/prompt-library/grammar-genie): Transforma oraciones gramaticalmente incorrectas en inglés apropiado.
- [Hal el ayudante humorístico](https://docs.anthropic.com/es/prompt-library/hal-the-humorous-helper): Chatea con una IA experta que tiene un lado sarcástico
- [Iluminador de modismos](https://docs.anthropic.com/es/prompt-library/idiom-illuminator): Explica el significado y origen de modismos y proverbios comunes.
- [Creador de preguntas para entrevistas](https://docs.anthropic.com/es/prompt-library/interview-question-crafter): Genera preguntas para entrevistas.
- [Leyenda LaTeX](https://docs.anthropic.com/es/prompt-library/latex-legend): Escriba documentos LaTeX, generando código para ecuaciones matemáticas, tablas y más.
- [Planificador de lecciones](https://docs.anthropic.com/es/prompt-library/lesson-planner): Crea planes de lecciones detallados sobre cualquier tema.
- [Biblioteca](https://docs.anthropic.com/es/prompt-library/library)
- [Moderador maestro](https://docs.anthropic.com/es/prompt-library/master-moderator): Evaluar las entradas del usuario en busca de contenido potencialmente dañino o ilegal.
- [Escriba de reuniones](https://docs.anthropic.com/es/prompt-library/meeting-scribe): Destila reuniones en resúmenes concisos que incluyen temas de discusión, puntos clave y elementos de acción.
- [Memo maestro](https://docs.anthropic.com/es/prompt-library/memo-maestro): Componer memorandos empresariales completos basados en puntos clave
- [Mentor de atención plena](https://docs.anthropic.com/es/prompt-library/mindfulness-mentor): Guía al usuario a través de ejercicios y técnicas de atención plena para reducir el estrés.
- [Colorante de estado de ánimo](https://docs.anthropic.com/es/prompt-library/mood-colorizer): Transforma descripciones textuales de estados de ánimo en códigos HEX correspondientes.
- [Musa motivacional](https://docs.anthropic.com/es/prompt-library/motivational-muse): Proporciona mensajes motivacionales y afirmaciones personalizadas basadas en la entrada del usuario
- [Creador de neologismos](https://docs.anthropic.com/es/prompt-library/neologism-creator): Inventa nuevas palabras y proporciona sus definiciones basadas en conceptos o ideas proporcionadas por el usuario.
- [Analizador de perspectivas](https://docs.anthropic.com/es/prompt-library/perspectives-ponderer): Evalúa los pros y contras de un tema proporcionado por el usuario.
- [Reflexiones filosóficas](https://docs.anthropic.com/es/prompt-library/philosophical-musings): Participa en discusiones filosóficas profundas y experimentos mentales.
- [Purificador de IIP](https://docs.anthropic.com/es/prompt-library/pii-purifier): Detecta y elimina automáticamente la información de identificación personal (IIP) de documentos de texto.
- [Superpoderes políglotas](https://docs.anthropic.com/es/prompt-library/polyglot-superpowers): Traduce texto de cualquier idioma a cualquier idioma.
- [Poeta de palabras combinadas](https://docs.anthropic.com/es/prompt-library/portmanteau-poet): Combina dos palabras para crear un nuevo portmanteau significativo.
- [Profesional de nombres de productos](https://docs.anthropic.com/es/prompt-library/product-naming-pro): Crea nombres atractivos de productos a partir de descripciones y palabras clave
- [Pulidor de prosa](https://docs.anthropic.com/es/prompt-library/prose-polisher): Refina y mejora el contenido escrito con técnicas avanzadas de corrección y sugerencias.
- [Pun-dit](https://docs.anthropic.com/es/prompt-library/pun-dit): Genera juegos de palabras e ingeniosos retruécanos basados en cualquier tema dado.
- [Eliminador de errores de Python](https://docs.anthropic.com/es/prompt-library/python-bug-buster): Detecta y corrige errores en código Python.
- [Clasificador de reseñas](https://docs.anthropic.com/es/prompt-library/review-classifier): Categorizar comentarios en etiquetas y categorizaciones predefinidas.
- [Adivina adivinanza](https://docs.anthropic.com/es/prompt-library/riddle-me-this): Genera adivinanzas y guía al usuario hacia las soluciones.
- [Simulador de escenarios de ciencia ficción](https://docs.anthropic.com/es/prompt-library/sci-fi-scenario-simulator): Discute con el usuario varios escenarios de ciencia ficción y los desafíos y consideraciones asociados.
- [Simplificador de segundo grado](https://docs.anthropic.com/es/prompt-library/second-grade-simplifier): Hacer que textos complejos sean fáciles de entender para estudiantes jóvenes
- [Experto en símiles](https://docs.anthropic.com/es/prompt-library/simile-savant): Genera símiles a partir de descripciones básicas
- [Sabio socrático](https://docs.anthropic.com/es/prompt-library/socratic-sage): Participa en una conversación de estilo socrático sobre un tema proporcionado por el usuario.
- [Mago de las hojas de cálculo](https://docs.anthropic.com/es/prompt-library/spreadsheet-sorcerer): Genera hojas de cálculo CSV con varios tipos de datos.
- [Hechicero SQL](https://docs.anthropic.com/es/prompt-library/sql-sorcerer): Transforma el lenguaje cotidiano en consultas SQL.
- [Compañero de narración](https://docs.anthropic.com/es/prompt-library/storytelling-sidekick): Crea historias cautivadoras de forma colaborativa con el usuario, ofreciendo giros argumentales y desarrollo de personajes.
- [Consultor de viajes en el tiempo](https://docs.anthropic.com/es/prompt-library/time-travel-consultant): Ayuda al usuario a navegar por escenarios hipotéticos de viajes en el tiempo y sus implicaciones.
- [Trabalenguas](https://docs.anthropic.com/es/prompt-library/tongue-twister): Crea trabalenguas desafiantes.
- [Generador de trivia](https://docs.anthropic.com/es/prompt-library/trivia-generator): Genera preguntas de trivia sobre una amplia gama de temas y proporciona pistas cuando sea necesario.
- [Detector de tono de tweets](https://docs.anthropic.com/es/prompt-library/tweet-tone-detector): Detecta el tono y sentimiento detrás de los tweets.
- [Innovador de fitness en RV](https://docs.anthropic.com/es/prompt-library/vr-fitness-innovator): Generar ideas creativas para juegos de fitness en realidad virtual.
- [Asistente de sitios web](https://docs.anthropic.com/es/prompt-library/website-wizard): Crea sitios web de una página basados en especificaciones del usuario.
- [API](https://docs.anthropic.com/es/release-notes/api): Sigue las actualizaciones de la API de Anthropic y la Consola de Desarrollador.
- [Aplicaciones de Claude](https://docs.anthropic.com/es/release-notes/claude-apps): Sigue las actualizaciones de las aplicaciones de Claude de Anthropic.
- [Resumen](https://docs.anthropic.com/es/release-notes/overview): Manténgase al día con las actualizaciones de los productos y servicios de Anthropic.
- [Prompts del Sistema](https://docs.anthropic.com/es/release-notes/system-prompts): Ver actualizaciones de los prompts del sistema principales en [Claude.ai](https://www.claude.ai) y las aplicaciones de Claude para [iOS](http://anthropic.com/ios) y [Android](http://anthropic.com/android).
- [Annuler un Lot de Messages (bêta)](https://docs.anthropic.com/fr/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [API Amazon Bedrock](https://docs.anthropic.com/fr/api/claude-on-amazon-bedrock): Les modèles Claude d'Anthropic sont maintenant généralement disponibles via Amazon Bedrock.
- [API Vertex AI](https://docs.anthropic.com/fr/api/claude-on-vertex-ai): Les modèles Claude d'Anthropic sont maintenant généralement disponibles via [Vertex AI](https://cloud.google.com/vertex-ai).
- [SDKs Client](https://docs.anthropic.com/fr/api/client-sdks): Nous fournissons des bibliothèques en Python et TypeScript qui facilitent le travail avec l'API Anthropic.
- [Créer un Lot de Messages (bêta)](https://docs.anthropic.com/fr/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Erreurs](https://docs.anthropic.com/fr/api/errors)
- [Obtenir de l'aide](https://docs.anthropic.com/fr/api/getting-help): Nous avons essayé de fournir les réponses aux questions les plus courantes dans cette documentation. Cependant, si vous avez besoin d'une assistance technique supplémentaire pour utiliser Claude, l'API Anthropic ou l'un de nos produits, vous pouvez contacter notre équipe d'assistance sur [support.anthropic.com](https://support.anthropic.com).
- [Pour commencer](https://docs.anthropic.com/fr/api/getting-started)
- [Adresses IP](https://docs.anthropic.com/fr/api/ip-addresses): Les services Anthropic sont hébergés sur une plage fixe d'adresses IP. Vous pouvez ajouter ces adresses à votre pare-feu pour ouvrir la surface minimale nécessaire au trafic sortant lors de l'accès à l'API et à la Console Anthropic. Ces plages ne changeront pas sans préavis.
- [Liste des Lots de Messages (bêta)](https://docs.anthropic.com/fr/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Exemples de lots de messages](https://docs.anthropic.com/fr/api/messages-batch-examples): Exemples d'utilisation de l'API Message Batches
- [Compter les tokens des messages (bêta)](https://docs.anthropic.com/fr/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Exemples de messages](https://docs.anthropic.com/fr/api/messages-examples): Exemples de requêtes et de réponses pour l'API Messages
- [Messages en Streaming](https://docs.anthropic.com/fr/api/messages-streaming)
- [Migration depuis les Text Completions](https://docs.anthropic.com/fr/api/migrating-from-text-completions-to-messages): Migration depuis les Text Completions vers les Messages
- [Validation des prompts](https://docs.anthropic.com/fr/api/prompt-validation): Avec les compléments de texte
- [Limites de débit](https://docs.anthropic.com/fr/api/rate-limits): Pour atténuer les abus et gérer la capacité de notre API, nous avons mis en place des limites sur l'utilisation de l'API Claude par une organisation.
- [Récupérer les Résultats du Lot de Messages (bêta)](https://docs.anthropic.com/fr/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Récupérer un Lot de Messages (bêta)](https://docs.anthropic.com/fr/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Diffusion de complétion de texte](https://docs.anthropic.com/fr/api/streaming)
- [Régions prises en charge](https://docs.anthropic.com/fr/api/supported-regions): Voici les pays, régions et territoires depuis lesquels nous pouvons actuellement prendre en charge l'accès :
- [Versions](https://docs.anthropic.com/fr/api/versioning): Lors des requêtes API, vous devez envoyer un en-tête de requête `anthropic-version`. Par exemple, `anthropic-version: 2023-06-01`. Si vous utilisez nos [bibliothèques client](/fr/api/client-libraries), cela est géré automatiquement pour vous.
- [Mises à jour de juillet 2024](https://docs.anthropic.com/fr/developer-newsletter/july2024): 28 juillet 2024
- [Vue d'ensemble](https://docs.anthropic.com/fr/developer-newsletter/overview): Explorez les mises à jour mensuelles, les analyses techniques approfondies, les meilleures pratiques et les réussites pour améliorer vos intégrations Claude
- [Modèles](https://docs.anthropic.com/fr/docs/about-claude/models): Claude est une famille de modèles de langage de pointe développés par Anthropic. Ce guide présente nos modèles et compare leurs performances avec les modèles existants.
- [Modération de contenu](https://docs.anthropic.com/fr/docs/about-claude/use-case-guides/content-moderation): La modération de contenu est un aspect essentiel du maintien d'un environnement sûr, respectueux et productif dans les applications numériques. Dans ce guide, nous verrons comment Claude peut être utilisé pour modérer le contenu au sein de votre application numérique.
- [Agent de service client](https://docs.anthropic.com/fr/docs/about-claude/use-case-guides/customer-support-chat): Ce guide explique comment exploiter les capacités conversationnelles avancées de Claude pour gérer les demandes des clients en temps réel, fournir un support 24/7, réduire les temps d'attente et gérer des volumes importants de support avec des réponses précises et des interactions positives.
- [Résumé juridique](https://docs.anthropic.com/fr/docs/about-claude/use-case-guides/legal-summarization): Ce guide explique comment exploiter les capacités avancées de traitement du langage naturel de Claude pour résumer efficacement les documents juridiques, en extrayant les informations clés et en accélérant la recherche juridique. Avec Claude, vous pouvez rationaliser l'examen des contrats, la préparation des litiges et le travail réglementaire, gagnant ainsi du temps et assurant la précision de vos processus juridiques.
- [Guides pour les cas d'utilisation courants](https://docs.anthropic.com/fr/docs/about-claude/use-case-guides/overview)
- [Routage des tickets](https://docs.anthropic.com/fr/docs/about-claude/use-case-guides/ticket-routing): Ce guide explique comment exploiter les capacités avancées de compréhension du langage naturel de Claude pour classifier les tickets de support client à grande échelle en fonction de l'intention du client, de l'urgence, de la priorisation, du profil client, et plus encore.
- [Extension Google Sheets](https://docs.anthropic.com/fr/docs/build-with-claude/claude-for-sheets): L'[extension Claude pour Sheets](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) intègre Claude dans Google Sheets, permettant d'exécuter des interactions avec Claude directement dans les cellules.
- [Utilisation de l'ordinateur (bêta)](https://docs.anthropic.com/fr/docs/build-with-claude/computer-use): Le modèle amélioré Claude 3.5 Sonnet est capable d'interagir avec des [outils](/fr/docs/build-with-claude/tool-use) qui peuvent manipuler un environnement de bureau informatique.
- [Définissez vos critères de réussite](https://docs.anthropic.com/fr/docs/build-with-claude/define-success)
- [Créer des évaluations empiriques solides](https://docs.anthropic.com/fr/docs/build-with-claude/develop-tests)
- [Plongements](https://docs.anthropic.com/fr/docs/build-with-claude/embeddings): Les plongements de texte sont des représentations numériques du texte qui permettent de mesurer la similarité sémantique. Ce guide présente les plongements, leurs applications et comment utiliser les modèles de plongement pour des tâches comme la recherche, les recommandations et la détection d'anomalies.
- [Lots de Messages (bêta)](https://docs.anthropic.com/fr/docs/build-with-claude/message-batches)
- [Support PDF (bêta)](https://docs.anthropic.com/fr/docs/build-with-claude/pdf-support): Les modèles Claude 3.5 Sonnet prennent désormais en charge l'entrée PDF et comprennent à la fois le contenu textuel et visuel des documents.
- [Mise en cache des prompts (bêta)](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-caching)
- [Soyez clair, direct et détaillé](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Laissez Claude réfléchir (prompting par chaîne de pensée) pour améliorer ses performances](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Chaîner des prompts complexes pour de meilleures performances](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Conseils pour les prompts à contexte long](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Utiliser des exemples (prompting multishot) pour guider le comportement de Claude](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Vue d'ensemble de l'ingénierie des prompts](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/overview)
- [Préremplir la réponse de Claude pour un meilleur contrôle des résultats](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Générer automatiquement des modèles de prompts en première version](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Utilisez notre améliorateur de prompts pour optimiser vos prompts](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Utiliser des modèles de prompts et des variables](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Donner un rôle à Claude avec une invite système](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/system-prompts)
- [Utiliser des balises XML pour structurer vos prompts](https://docs.anthropic.com/fr/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Génération de texte](https://docs.anthropic.com/fr/docs/build-with-claude/text-generation): Claude excelle dans un large éventail de tâches textuelles. Claude a été entraîné pour ingérer du code, de la prose et d'autres entrées en langage naturel, et fournir des sorties textuelles en réponse.
- [Comptage des tokens (bêta)](https://docs.anthropic.com/fr/docs/build-with-claude/token-counting)
- [Utilisation d'outils (appel de fonctions)](https://docs.anthropic.com/fr/docs/build-with-claude/tool-use)
- [Vision](https://docs.anthropic.com/fr/docs/build-with-claude/vision): La famille de modèles Claude 3 est dotée de nouvelles capacités de vision qui permettent à Claude de comprendre et d'analyser des images, ouvrant ainsi des possibilités passionnantes d'interaction multimodale.
- [Configuration initiale](https://docs.anthropic.com/fr/docs/initial-setup): Apprenons à utiliser l'API Anthropic pour construire avec Claude.
- [Introduction à Claude](https://docs.anthropic.com/fr/docs/intro-to-claude): Claude est une famille de [modèles d'IA hautement performants et intelligents](/fr/docs/about-claude/models) créée par Anthropic. Bien que Claude soit puissant et extensible, c'est aussi l'IA la plus fiable et la plus sûre disponible. Il suit des protocoles critiques, fait moins d'erreurs et résiste aux tentatives de contournement—permettant aux [clients entreprises](https://www.anthropic.com/customers) de créer les applications alimentées par l'IA les plus sûres à grande échelle.
- [Politique de Confidentialité d'Anthropic](https://docs.anthropic.com/fr/docs/legal-center/privacy)
- [Guide Anthropic](https://docs.anthropic.com/fr/docs/resources/cookbook)
- [Cours Anthropic](https://docs.anthropic.com/fr/docs/resources/courses)
- [Glossaire](https://docs.anthropic.com/fr/docs/resources/glossary): Ces concepts ne sont pas uniques aux modèles de langage d'Anthropic, mais nous présentons ci-dessous un bref résumé des termes clés.
- [Fiche technique du modèle Claude 3](https://docs.anthropic.com/fr/docs/resources/model-card)
- [Obsolescence des modèles](https://docs.anthropic.com/fr/docs/resources/model-deprecations)
- [État du système](https://docs.anthropic.com/fr/docs/resources/status)
- [Utilisation de l'outil d'évaluation](https://docs.anthropic.com/fr/docs/test-and-evaluate/eval-tool): La [Console Anthropic](https://console.anthropic.com/dashboard) propose un **outil d'évaluation** qui vous permet de tester vos prompts dans différents scénarios.
- [Améliorer la cohérence des réponses (mode JSON)](https://docs.anthropic.com/fr/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Garder Claude dans son rôle avec le prompt de rôle et le pré-remplissage](https://docs.anthropic.com/fr/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Atténuer les contournements et les injections de prompts](https://docs.anthropic.com/fr/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Réduire les hallucinations](https://docs.anthropic.com/fr/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Réduction de la latence](https://docs.anthropic.com/fr/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Réduire la fuite de prompt](https://docs.anthropic.com/fr/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Bienvenue à Claude](https://docs.anthropic.com/fr/docs/welcome): Claude est une plateforme d'IA hautement performante, fiable et intelligente créée par Anthropic. Claude excelle dans les tâches impliquant le langage, le raisonnement, l'analyse, la programmation et plus encore.
- [null](https://docs.anthropic.com/fr/home)
- [Éditeur adaptatif](https://docs.anthropic.com/fr/prompt-library/adaptive-editor): Réécrivez le texte selon les instructions données par l'utilisateur, comme avec un ton, un public ou un style différent.
- [Analyseur de codes d'aéroport](https://docs.anthropic.com/fr/prompt-library/airport-code-analyst): Trouver et extraire les codes d'aéroport du texte.
- [Anthropologue extraterrestre](https://docs.anthropic.com/fr/prompt-library/alien-anthropologist): Analyser la culture et les coutumes humaines du point de vue d'un anthropologue extraterrestre.
- [Alchimiste de l'allitération](https://docs.anthropic.com/fr/prompt-library/alliteration-alchemist): Générer des phrases et des expressions allitératives pour n'importe quel sujet.
- [Les diffusions de Babel](https://docs.anthropic.com/fr/prompt-library/babels-broadcasts): Créez des tweets d'annonce de produits captivants dans les 10 langues les plus parlées au monde.
- [Créateur de marque](https://docs.anthropic.com/fr/prompt-library/brand-builder): Élaborer un brief de conception pour une identité de marque holistique.
- [Coach de carrière](https://docs.anthropic.com/fr/prompt-library/career-coach): Engagez des conversations en jeu de rôle avec un coach de carrière IA.
- [Citez vos sources](https://docs.anthropic.com/fr/prompt-library/cite-your-sources): Obtenez des réponses aux questions sur le contenu d'un document avec des citations pertinentes à l'appui de la réponse.
- [Clarificateur de code](https://docs.anthropic.com/fr/prompt-library/code-clarifier): Simplifier et expliquer du code complexe en langage simple.
- [Consultant en code](https://docs.anthropic.com/fr/prompt-library/code-consultant): Suggérer des améliorations pour optimiser les performances du code Python.
- [Clairvoyant d'entreprise](https://docs.anthropic.com/fr/prompt-library/corporate-clairvoyant): Extraire des informations, identifier les risques et distiller les informations clés des longs rapports d'entreprise en une seule note.
- [Frappes Cosmiques](https://docs.anthropic.com/fr/prompt-library/cosmic-keystrokes): Générez un jeu interactif de frappe rapide dans un seul fichier HTML, avec un gameplay à défilement latéral et un style Tailwind CSS.
- [Convertisseur CSV](https://docs.anthropic.com/fr/prompt-library/csv-converter): Convertissez des données de différents formats (JSON, XML, etc.) en fichiers CSV correctement formatés.
- [Créateur culinaire](https://docs.anthropic.com/fr/prompt-library/culinary-creator): Suggérer des idées de recettes en fonction des ingrédients disponibles et des préférences alimentaires de l'utilisateur.
- [Organisateur de données](https://docs.anthropic.com/fr/prompt-library/data-organizer): Transformez du texte non structuré en tableaux JSON personnalisés.
- [Décodeur de directions](https://docs.anthropic.com/fr/prompt-library/direction-decoder): Transformez le langage naturel en instructions étape par étape.
- [Interprète de rêves](https://docs.anthropic.com/fr/prompt-library/dream-interpreter): Offrir des interprétations et des aperçus sur le symbolisme des rêves de l'utilisateur
- [Estimateur d'efficacité](https://docs.anthropic.com/fr/prompt-library/efficiency-estimator): Calculer la complexité temporelle des fonctions et des algorithmes.
- [Extracteur d'e-mails](https://docs.anthropic.com/fr/prompt-library/email-extractor): Extrait les adresses e-mail d'un document dans une liste formatée en JSON.
- [Encodeur d'émojis](https://docs.anthropic.com/fr/prompt-library/emoji-encoder): Convertissez du texte simple en messages amusants et expressifs avec des émojis.
- [Navigateur de dilemmes éthiques](https://docs.anthropic.com/fr/prompt-library/ethical-dilemma-navigator): Aide l'utilisateur à réfléchir à des dilemmes éthiques complexes et fournit différentes perspectives.
- [Expert en formules Excel](https://docs.anthropic.com/fr/prompt-library/excel-formula-expert): Créer des formules Excel basées sur des calculs ou des manipulations de données décrits par l'utilisateur.
- [Fabricant de fonctions](https://docs.anthropic.com/fr/prompt-library/function-fabricator): Créez des fonctions Python basées sur des spécifications détaillées.
- [Conseiller en mode futuriste](https://docs.anthropic.com/fr/prompt-library/futuristic-fashion-advisor): Suggère des tendances et styles de mode avant-gardistes selon les préférences spécifiques de l'utilisateur
- [Maîtriser Git](https://docs.anthropic.com/fr/prompt-library/git-gud): Générer des commandes Git appropriées basées sur les actions de contrôle de version décrites par l'utilisateur.
- [Scripteur Google apps](https://docs.anthropic.com/fr/prompt-library/google-apps-scripter): Générer des scripts Google Apps pour accomplir des tâches selon les besoins de l'utilisateur.
- [Guru de l'évaluation](https://docs.anthropic.com/fr/prompt-library/grading-guru): Comparez et évaluez la qualité des textes écrits selon des critères et des normes définis par l'utilisateur.
- [Génie de la grammaire](https://docs.anthropic.com/fr/prompt-library/grammar-genie): Transformez des phrases grammaticalement incorrectes en anglais correct.
- [Hal l'assistant humoristique](https://docs.anthropic.com/fr/prompt-library/hal-the-humorous-helper): Discutez avec une IA érudite qui a un côté sarcastique.
- [Illuminateur d'expressions idiomatiques](https://docs.anthropic.com/fr/prompt-library/idiom-illuminator): Expliquer le sens et l'origine des expressions idiomatiques et proverbes courants
- [Générateur de questions d'entretien](https://docs.anthropic.com/fr/prompt-library/interview-question-crafter): Générer des questions pour les entretiens.
- [Légende LaTeX](https://docs.anthropic.com/fr/prompt-library/latex-legend): Rédigez des documents LaTeX, en générant du code pour des équations mathématiques, des tableaux et plus encore.
- [Planificateur de leçons](https://docs.anthropic.com/fr/prompt-library/lesson-planner): Créez des plans de leçons approfondis sur n'importe quel sujet.
- [Bibliothèque](https://docs.anthropic.com/fr/prompt-library/library)
- [Modérateur expert](https://docs.anthropic.com/fr/prompt-library/master-moderator): Évaluer les entrées des utilisateurs pour détecter les contenus potentiellement dangereux ou illégaux.
- [Scribe de réunion](https://docs.anthropic.com/fr/prompt-library/meeting-scribe): Distillez les réunions en résumés concis incluant les sujets de discussion, les points clés et les actions à entreprendre.
- [Maître des mémos](https://docs.anthropic.com/fr/prompt-library/memo-maestro): Rédigez des mémos d'entreprise complets basés sur des points clés.
- [Mentor de pleine conscience](https://docs.anthropic.com/fr/prompt-library/mindfulness-mentor): Guidez l'utilisateur à travers des exercices de pleine conscience et des techniques de réduction du stress.
- [Coloriseur d'humeur](https://docs.anthropic.com/fr/prompt-library/mood-colorizer): Transforme les descriptions textuelles d'humeurs en codes HEX correspondants.
- [Muse motivationnelle](https://docs.anthropic.com/fr/prompt-library/motivational-muse): Fournir des messages motivants et des affirmations personnalisés basés sur les contributions de l'utilisateur
- [Créateur de néologismes](https://docs.anthropic.com/fr/prompt-library/neologism-creator): Inventez de nouveaux mots et fournissez leurs définitions basées sur les concepts ou idées fournis par l'utilisateur.
- [Évaluateur de perspectives](https://docs.anthropic.com/fr/prompt-library/perspectives-ponderer): Évaluer les avantages et les inconvénients d'un sujet fourni par l'utilisateur.
- [Réflexions philosophiques](https://docs.anthropic.com/fr/prompt-library/philosophical-musings): S'engager dans des discussions philosophiques profondes et des expériences de pensée.
- [Purificateur de PII](https://docs.anthropic.com/fr/prompt-library/pii-purifier): Détectez et supprimez automatiquement les informations personnelles identifiables (PII) des documents texte.
- [Superpouvoirs polyglottes](https://docs.anthropic.com/fr/prompt-library/polyglot-superpowers): Traduisez du texte de n'importe quelle langue vers n'importe quelle langue.
- [Poète des mots-valises](https://docs.anthropic.com/fr/prompt-library/portmanteau-poet): Fusionnez deux mots pour créer un nouveau mot-valise significatif.
- [Pro des noms de produits](https://docs.anthropic.com/fr/prompt-library/product-naming-pro): Créez des noms de produits accrocheurs à partir de descriptions et de mots-clés.
- [Polisseur de prose](https://docs.anthropic.com/fr/prompt-library/prose-polisher): Affinez et améliorez le contenu écrit avec des techniques avancées de révision et des suggestions.
- [Jeu de mots](https://docs.anthropic.com/fr/prompt-library/pun-dit): Générer des jeux de mots et des calembours intelligents sur n'importe quel sujet donné.
- [Débogueur Python](https://docs.anthropic.com/fr/prompt-library/python-bug-buster): Détecter et corriger les bogues dans le code Python.
- [Classificateur d'avis](https://docs.anthropic.com/fr/prompt-library/review-classifier): Catégoriser les retours d'information en étiquettes et catégorisations prédéfinies.
- [Posez-moi une énigme](https://docs.anthropic.com/fr/prompt-library/riddle-me-this): Générer des énigmes et guider l'utilisateur vers les solutions.
- [Simulateur de scénarios de science-fiction](https://docs.anthropic.com/fr/prompt-library/sci-fi-scenario-simulator): Discuter avec l'utilisateur de divers scénarios de science-fiction et des défis et considérations associés.
- [Simplificateur de niveau CE1](https://docs.anthropic.com/fr/prompt-library/second-grade-simplifier): Rendre les textes complexes faciles à comprendre pour les jeunes apprenants
- [Expert en comparaisons](https://docs.anthropic.com/fr/prompt-library/simile-savant): Générer des comparaisons à partir de descriptions simples.
- [Sage socratique](https://docs.anthropic.com/fr/prompt-library/socratic-sage): Engagez une conversation de style socratique sur un sujet donné par l'utilisateur.
- [Magicien des feuilles de calcul](https://docs.anthropic.com/fr/prompt-library/spreadsheet-sorcerer): Générer des feuilles de calcul CSV avec différents types de données.
- [Sorcier SQL](https://docs.anthropic.com/fr/prompt-library/sql-sorcerer): Transformez le langage courant en requêtes SQL.
- [Assistant de narration](https://docs.anthropic.com/fr/prompt-library/storytelling-sidekick): Créez des histoires captivantes en collaboration avec l'utilisateur, en proposant des rebondissements et un développement des personnages.
- [Consultant en voyage dans le temps](https://docs.anthropic.com/fr/prompt-library/time-travel-consultant): Aide l'utilisateur à naviguer dans des scénarios hypothétiques de voyage dans le temps et leurs implications.
- [Virelangues](https://docs.anthropic.com/fr/prompt-library/tongue-twister): Créer des virelangues stimulants.
- [Générateur de quiz](https://docs.anthropic.com/fr/prompt-library/trivia-generator): Générez des questions de quiz sur un large éventail de sujets et fournissez des indices si nécessaire.
- [Détecteur de ton des tweets](https://docs.anthropic.com/fr/prompt-library/tweet-tone-detector): Détectez le ton et le sentiment derrière les tweets.
- [Innovateur en fitness VR](https://docs.anthropic.com/fr/prompt-library/vr-fitness-innovator): Réfléchir à des idées créatives pour des jeux de fitness en réalité virtuelle.
- [Assistant de création de site web](https://docs.anthropic.com/fr/prompt-library/website-wizard): Créez des sites web d'une page selon les spécifications de l'utilisateur.
- [API](https://docs.anthropic.com/fr/release-notes/api): Suivez les mises à jour de l'API Anthropic et de la Console Développeur.
- [Applications Claude](https://docs.anthropic.com/fr/release-notes/claude-apps): Suivez les mises à jour des applications Claude d'Anthropic.
- [Vue d'ensemble](https://docs.anthropic.com/fr/release-notes/overview): Suivez les mises à jour des produits et services d'Anthropic.
- [Prompts Système](https://docs.anthropic.com/fr/release-notes/system-prompts): Consultez les mises à jour des prompts système principaux sur [Claude.ai](https://www.claude.ai) et les applications Claude [iOS](http://anthropic.com/ios) et [Android](http://anthropic.com/android).
- [Membatalkan Batch Pesan (beta)](https://docs.anthropic.com/id/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [API Amazon Bedrock](https://docs.anthropic.com/id/api/claude-on-amazon-bedrock): Model Claude dari Anthropic kini tersedia secara umum melalui Amazon Bedrock.
- [Vertex AI API](https://docs.anthropic.com/id/api/claude-on-vertex-ai): Model Claude dari Anthropic sekarang tersedia secara umum melalui [Vertex AI](https://cloud.google.com/vertex-ai).
- [SDK Klien](https://docs.anthropic.com/id/api/client-sdks): Kami menyediakan pustaka dalam Python dan TypeScript yang memudahkan penggunaan API Anthropic.
- [Create a Text Completion](https://docs.anthropic.com/id/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Membuat Batch Pesan (beta)](https://docs.anthropic.com/id/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Error](https://docs.anthropic.com/id/api/errors)
- [Mendapatkan bantuan](https://docs.anthropic.com/id/api/getting-help): Kami telah berusaha menyediakan jawaban untuk pertanyaan yang paling umum dalam dokumentasi ini. Namun, jika Anda membutuhkan dukungan teknis lebih lanjut dalam menggunakan Claude, API Anthropic, atau produk kami lainnya, Anda dapat menghubungi tim dukungan kami di [support.anthropic.com](https://support.anthropic.com).
- [Memulai](https://docs.anthropic.com/id/api/getting-started)
- [Alamat IP](https://docs.anthropic.com/id/api/ip-addresses): Layanan Anthropic berada pada rentang alamat IP tetap. Anda dapat menambahkan ini ke firewall Anda untuk membuka area permukaan minimum untuk lalu lintas keluar saat mengakses API dan Konsol Anthropic. Rentang ini tidak akan berubah tanpa pemberitahuan.
- [Daftar Batch Pesan (beta)](https://docs.anthropic.com/id/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/id/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Contoh Batch Pesan](https://docs.anthropic.com/id/api/messages-batch-examples): Contoh penggunaan API Batch Pesan
- [Menghitung token Pesan (beta)](https://docs.anthropic.com/id/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Contoh pesan](https://docs.anthropic.com/id/api/messages-examples): Contoh permintaan dan respons untuk Messages API
- [Streaming Pesan](https://docs.anthropic.com/id/api/messages-streaming)
- [Migrasi dari Text Completions](https://docs.anthropic.com/id/api/migrating-from-text-completions-to-messages): Migrasi dari Text Completions ke Messages
- [Validasi prompt](https://docs.anthropic.com/id/api/prompt-validation): Dengan Text Completions
- [Batas penggunaan](https://docs.anthropic.com/id/api/rate-limits): Untuk mengurangi penyalahgunaan dan mengelola kapasitas API kami, kami telah menerapkan batasan pada seberapa banyak sebuah organisasi dapat menggunakan API Claude.
- [Mengambil Hasil Batch Pesan (beta)](https://docs.anthropic.com/id/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Mengambil Batch Pesan (beta)](https://docs.anthropic.com/id/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Streaming Text Completions](https://docs.anthropic.com/id/api/streaming)
- [Wilayah yang didukung](https://docs.anthropic.com/id/api/supported-regions): Berikut adalah negara, wilayah, dan teritori yang saat ini dapat kami dukung aksesnya:
- [Versi](https://docs.anthropic.com/id/api/versioning): Saat membuat permintaan API, Anda harus mengirimkan header permintaan `anthropic-version`. Contohnya, `anthropic-version: 2023-06-01`. Jika Anda menggunakan [pustaka klien](/id/api/client-libraries) kami, ini akan ditangani secara otomatis untuk Anda.
- [Pembaruan Juli 2024](https://docs.anthropic.com/id/developer-newsletter/july2024): 28 Juli 2024
- [Ikhtisar](https://docs.anthropic.com/id/developer-newsletter/overview): Jelajahi pembaruan bulanan, pembahasan mendalam teknis, praktik terbaik, dan kisah sukses untuk meningkatkan integrasi Claude Anda
- [Model](https://docs.anthropic.com/id/docs/about-claude/models): Claude adalah keluarga model bahasa besar mutakhir yang dikembangkan oleh Anthropic. Panduan ini memperkenalkan model kami dan membandingkan kinerjanya dengan model lama.
- [Keamanan dan kepatuhan](https://docs.anthropic.com/id/docs/about-claude/security-compliance)
- [Moderasi konten](https://docs.anthropic.com/id/docs/about-claude/use-case-guides/content-moderation): Moderasi konten adalah aspek penting dalam menjaga lingkungan yang aman, hormat, dan produktif dalam aplikasi digital. Dalam panduan ini, kita akan membahas bagaimana Claude dapat digunakan untuk memoderasi konten dalam aplikasi digital Anda.
- [Agen dukungan pelanggan](https://docs.anthropic.com/id/docs/about-claude/use-case-guides/customer-support-chat): Panduan ini menjelaskan cara memanfaatkan kemampuan percakapan Claude yang canggih untuk menangani pertanyaan pelanggan secara real-time, menyediakan dukungan 24/7, mengurangi waktu tunggu, dan mengelola volume dukungan yang tinggi dengan respons akurat dan interaksi positif.
- [Peringkasan dokumen hukum](https://docs.anthropic.com/id/docs/about-claude/use-case-guides/legal-summarization): Panduan ini menjelaskan cara memanfaatkan kemampuan pemrosesan bahasa alami tingkat lanjut Claude untuk meringkas dokumen hukum secara efisien, mengekstrak informasi kunci dan mempercepat penelitian hukum. Dengan Claude, Anda dapat merampingkan peninjauan kontrak, persiapan litigasi, dan pekerjaan regulasi, menghemat waktu dan memastikan akurasi dalam proses hukum Anda.
- [Panduan untuk kasus penggunaan umum](https://docs.anthropic.com/id/docs/about-claude/use-case-guides/overview)
- [Perutean tiket](https://docs.anthropic.com/id/docs/about-claude/use-case-guides/ticket-routing): Panduan ini menjelaskan cara memanfaatkan kemampuan pemahaman bahasa alami tingkat lanjut Claude untuk mengklasifikasikan tiket dukungan pelanggan secara massal berdasarkan maksud pelanggan, urgensi, prioritas, profil pelanggan, dan lainnya.
- [Add-on Google Sheets](https://docs.anthropic.com/id/docs/build-with-claude/claude-for-sheets): [Ekstensi Claude for Sheets](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) mengintegrasikan Claude ke dalam Google Sheets, memungkinkan Anda untuk menjalankan interaksi dengan Claude langsung di dalam sel.
- [Penggunaan Komputer (beta)](https://docs.anthropic.com/id/docs/build-with-claude/computer-use): Model Claude 3.5 Sonnet yang ditingkatkan mampu berinteraksi dengan [alat](/id/docs/build-with-claude/tool-use) yang dapat memanipulasi lingkungan desktop komputer.
- [Tentukan kriteria keberhasilan Anda](https://docs.anthropic.com/id/docs/build-with-claude/define-success)
- [Membuat evaluasi empiris yang kuat](https://docs.anthropic.com/id/docs/build-with-claude/develop-tests)
- [Embeddings](https://docs.anthropic.com/id/docs/build-with-claude/embeddings): Embeddings teks adalah representasi numerik dari teks yang memungkinkan pengukuran kesamaan semantik. Panduan ini memperkenalkan embeddings, aplikasinya, dan cara menggunakan model embedding untuk tugas-tugas seperti pencarian, rekomendasi, dan deteksi anomali.
- [Batch Pesan (beta)](https://docs.anthropic.com/id/docs/build-with-claude/message-batches)
- [Dukungan PDF (beta)](https://docs.anthropic.com/id/docs/build-with-claude/pdf-support): Model Claude 3.5 Sonnet sekarang mendukung input PDF dan memahami konten teks dan visual dalam dokumen.
- [Penyimpanan cache prompt (beta)](https://docs.anthropic.com/id/docs/build-with-claude/prompt-caching)
- [Bersikaplah jelas, langsung, dan terperinci](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Biarkan Claude berpikir (chain of thought prompting) untuk meningkatkan performa](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Merangkai prompt kompleks untuk performa yang lebih kuat](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Tips untuk prompt konteks panjang](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Gunakan contoh (multishot prompting) untuk memandu perilaku Claude](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Gambaran umum rekayasa prompt](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/overview)
- [Mengisi awal respons Claude untuk kontrol output yang lebih baik](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Secara otomatis menghasilkan template prompt draf pertama](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Gunakan penyempurna prompt kami untuk mengoptimalkan prompt Anda](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Menggunakan template prompt dan variabel](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Memberikan Claude sebuah peran dengan prompt sistem](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/system-prompts)
- [Gunakan tag XML untuk menyusun prompt Anda](https://docs.anthropic.com/id/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Pembuatan teks](https://docs.anthropic.com/id/docs/build-with-claude/text-generation): Claude unggul dalam berbagai tugas berbasis teks. Claude telah dilatih untuk memproses kode, prosa, dan input bahasa alami lainnya, serta memberikan output teks sebagai respons.
- [Penghitungan token (beta)](https://docs.anthropic.com/id/docs/build-with-claude/token-counting)
- [Penggunaan alat (pemanggilan fungsi)](https://docs.anthropic.com/id/docs/build-with-claude/tool-use)
- [Visi](https://docs.anthropic.com/id/docs/build-with-claude/vision): Keluarga model Claude 3 hadir dengan kemampuan visi baru yang memungkinkan Claude untuk memahami dan menganalisis gambar, membuka kemungkinan menarik untuk interaksi multimodal.
- [Pengaturan awal](https://docs.anthropic.com/id/docs/initial-setup): Mari pelajari cara menggunakan API Anthropic untuk membangun dengan Claude.
- [Pengantar Claude](https://docs.anthropic.com/id/docs/intro-to-claude): Claude adalah keluarga [model AI yang sangat berkinerja dan cerdas](/id/docs/about-claude/models) yang dibuat oleh Anthropic. Meskipun Claude kuat dan dapat diperluas, ini juga merupakan AI yang paling dapat dipercaya dan andal yang tersedia. Ia mengikuti protokol penting, membuat lebih sedikit kesalahan, dan tahan terhadap jailbreak—memungkinkan [pelanggan enterprise](https://www.anthropic.com/customers) untuk membangun aplikasi bertenaga AI yang paling aman dalam skala besar.
- [Kebijakan Privasi Anthropic](https://docs.anthropic.com/id/docs/legal-center/privacy)
- [Anthropic Cookbook](https://docs.anthropic.com/id/docs/resources/cookbook)
- [Kursus Anthropic](https://docs.anthropic.com/id/docs/resources/courses)
- [Glosarium](https://docs.anthropic.com/id/docs/resources/glossary): Konsep-konsep ini tidak unik untuk model bahasa Anthropic, tetapi kami menyajikan ringkasan singkat istilah-istilah kunci di bawah ini.
- [Kartu model Claude 3](https://docs.anthropic.com/id/docs/resources/model-card)
- [Penghentian Model](https://docs.anthropic.com/id/docs/resources/model-deprecations)
- [Status sistem](https://docs.anthropic.com/id/docs/resources/status)
- [Menggunakan Alat Evaluasi](https://docs.anthropic.com/id/docs/test-and-evaluate/eval-tool): [Anthropic Console](https://console.anthropic.com/dashboard) memiliki **alat Evaluasi** yang memungkinkan Anda menguji prompt Anda dalam berbagai skenario.
- [Meningkatkan konsistensi output (mode JSON)](https://docs.anthropic.com/id/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Menjaga Claude tetap dalam karakter dengan role prompting dan prefilling](https://docs.anthropic.com/id/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Mengatasi jailbreak dan injeksi prompt](https://docs.anthropic.com/id/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Mengurangi halusinasi](https://docs.anthropic.com/id/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Mengurangi latensi](https://docs.anthropic.com/id/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Mengurangi kebocoran prompt](https://docs.anthropic.com/id/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Selamat Datang di Claude](https://docs.anthropic.com/id/docs/welcome): Claude adalah platform AI yang sangat berkinerja, terpercaya, dan cerdas yang dibangun oleh Anthropic. Claude unggul dalam tugas-tugas yang melibatkan bahasa, penalaran, analisis, pemrograman, dan lainnya.
- [null](https://docs.anthropic.com/id/home)
- [Editor adaptif](https://docs.anthropic.com/id/prompt-library/adaptive-editor): Menulis ulang teks mengikuti instruksi yang diberikan pengguna, seperti dengan nada, audiens, atau gaya yang berbeda.
- [Analis kode bandara](https://docs.anthropic.com/id/prompt-library/airport-code-analyst): Menemukan dan mengekstrak kode bandara dari teks.
- [Antropolog alien](https://docs.anthropic.com/id/prompt-library/alien-anthropologist): Menganalisis budaya dan kebiasaan manusia dari perspektif seorang antropolog alien.
- [Alkemis aliterasi](https://docs.anthropic.com/id/prompt-library/alliteration-alchemist): Menghasilkan frasa dan kalimat aliterasi untuk subjek apapun.
- [Siaran Babel](https://docs.anthropic.com/id/prompt-library/babels-broadcasts): Buat tweet pengumuman produk yang menarik dalam 10 bahasa yang paling banyak digunakan di dunia.
- [Pembuat merek](https://docs.anthropic.com/id/prompt-library/brand-builder): Buat sebuah brief desain untuk identitas merek yang holistik.
- [Pelatih karir](https://docs.anthropic.com/id/prompt-library/career-coach): Lakukan percakapan bermain peran dengan pelatih karir AI.
- [Kutip sumber Anda](https://docs.anthropic.com/id/prompt-library/cite-your-sources): Dapatkan jawaban untuk pertanyaan tentang konten dokumen dengan kutipan relevan yang mendukung respons.
- [Penjelas kode](https://docs.anthropic.com/id/prompt-library/code-clarifier): Menyederhanakan dan menjelaskan kode yang kompleks dalam bahasa yang mudah dipahami.
- [Konsultan kode](https://docs.anthropic.com/id/prompt-library/code-consultant): Menyarankan perbaikan untuk mengoptimalkan kinerja kode Python.
- [Peramal korporat](https://docs.anthropic.com/id/prompt-library/corporate-clairvoyant): Ekstrak wawasan, identifikasi risiko, dan saring informasi kunci dari laporan korporat yang panjang menjadi satu memo.
- [Ketukan Kosmik](https://docs.anthropic.com/id/prompt-library/cosmic-keystrokes): Menghasilkan permainan mengetik interaktif dalam satu file HTML, dengan gameplay bergulir ke samping dan gaya Tailwind CSS.
- [Konverter CSV](https://docs.anthropic.com/id/prompt-library/csv-converter): Mengkonversi data dari berbagai format (JSON, XML, dll.) menjadi file CSV yang terformat dengan benar.
- [Kreator kuliner](https://docs.anthropic.com/id/prompt-library/culinary-creator): Menyarankan ide resep berdasarkan bahan-bahan yang tersedia dan preferensi makanan pengguna.
- [Pengorganisir Data](https://docs.anthropic.com/id/prompt-library/data-organizer): Mengubah teks tidak terstruktur menjadi tabel JSON yang disesuaikan.
- [Pengurai arah](https://docs.anthropic.com/id/prompt-library/direction-decoder): Mengubah bahasa alami menjadi petunjuk langkah demi langkah.
- [Penerjemah mimpi](https://docs.anthropic.com/id/prompt-library/dream-interpreter): Memberikan interpretasi dan wawasan tentang simbolisme dari mimpi pengguna
- [Penghitung efisiensi](https://docs.anthropic.com/id/prompt-library/efficiency-estimator): Menghitung kompleksitas waktu dari fungsi dan algoritma.
- [Pengekstrak email](https://docs.anthropic.com/id/prompt-library/email-extractor): Ekstrak alamat email dari dokumen menjadi daftar berformat JSON.
- [Penyandi emoji](https://docs.anthropic.com/id/prompt-library/emoji-encoder): Mengubah teks biasa menjadi pesan emoji yang menyenangkan dan ekspresif.
- [Navigator dilema etis](https://docs.anthropic.com/id/prompt-library/ethical-dilemma-navigator): Membantu pengguna memikirkan dilema etis yang kompleks dan memberikan berbagai perspektif.
- [Ahli rumus Excel](https://docs.anthropic.com/id/prompt-library/excel-formula-expert): Membuat rumus Excel berdasarkan perhitungan atau manipulasi data yang dijelaskan pengguna.
- [Pembuat fungsi](https://docs.anthropic.com/id/prompt-library/function-fabricator): Membuat fungsi Python berdasarkan spesifikasi yang terperinci.
- [Penasihat mode futuristik](https://docs.anthropic.com/id/prompt-library/futuristic-fashion-advisor): Menyarankan tren dan gaya mode avant-garde sesuai preferensi spesifik pengguna
- [Mahir Git](https://docs.anthropic.com/id/prompt-library/git-gud): Menghasilkan perintah Git yang sesuai berdasarkan tindakan kontrol versi yang dijelaskan pengguna.
- [Penulis skrip Google apps](https://docs.anthropic.com/id/prompt-library/google-apps-scripter): Menghasilkan skrip Google Apps untuk menyelesaikan tugas berdasarkan kebutuhan pengguna.
- [Guru penilaian](https://docs.anthropic.com/id/prompt-library/grading-guru): Membandingkan dan mengevaluasi kualitas teks tertulis berdasarkan kriteria dan standar yang ditentukan pengguna.
- [Penyihir tata bahasa](https://docs.anthropic.com/id/prompt-library/grammar-genie): Mengubah kalimat yang tidak benar secara tata bahasa menjadi bahasa Inggris yang tepat.
- [Hal si pembantu humoris](https://docs.anthropic.com/id/prompt-library/hal-the-humorous-helper): Mengobrol dengan AI berpengetahuan luas yang memiliki sisi sarkastik.
- [Penjelas Idiom](https://docs.anthropic.com/id/prompt-library/idiom-illuminator): Menjelaskan arti dan asal-usul idiom dan peribahasa yang umum.
- [Pembuat pertanyaan wawancara](https://docs.anthropic.com/id/prompt-library/interview-question-crafter): Menghasilkan pertanyaan untuk wawancara.
- [Legenda LaTeX](https://docs.anthropic.com/id/prompt-library/latex-legend): Tulis dokumen LaTeX, menghasilkan kode untuk persamaan matematika, tabel, dan lainnya.
- [Pembuat rencana pelajaran](https://docs.anthropic.com/id/prompt-library/lesson-planner): Buat rencana pelajaran mendalam untuk berbagai mata pelajaran.
- [Perpustakaan](https://docs.anthropic.com/id/prompt-library/library)
- [Moderator ahli](https://docs.anthropic.com/id/prompt-library/master-moderator): Mengevaluasi input pengguna untuk konten yang berpotensi berbahaya atau ilegal.
- [Pencatat rapat](https://docs.anthropic.com/id/prompt-library/meeting-scribe): Menyarikan rapat menjadi ringkasan singkat termasuk topik diskusi, poin-poin penting, dan item tindakan.
- [Maestro memo](https://docs.anthropic.com/id/prompt-library/memo-maestro): Menyusun memo perusahaan yang komprehensif berdasarkan poin-poin utama.
- [Mentor kesadaran penuh](https://docs.anthropic.com/id/prompt-library/mindfulness-mentor): Membimbing pengguna melalui latihan kesadaran penuh dan teknik untuk mengurangi stres.
- [Pewarna suasana hati](https://docs.anthropic.com/id/prompt-library/mood-colorizer): Mengubah deskripsi teks suasana hati menjadi kode HEX yang sesuai.
- [Muse Motivasi](https://docs.anthropic.com/id/prompt-library/motivational-muse): Memberikan pesan motivasi dan afirmasi yang dipersonalisasi berdasarkan masukan pengguna.
- [Pembuat kata baru](https://docs.anthropic.com/id/prompt-library/neologism-creator): Menciptakan kata-kata baru dan memberikan definisinya berdasarkan konsep atau ide yang diberikan pengguna.
- [Pemikir perspektif](https://docs.anthropic.com/id/prompt-library/perspectives-ponderer): Menimbang kelebihan dan kekurangan dari topik yang diberikan pengguna.
- [Perenungan filosofis](https://docs.anthropic.com/id/prompt-library/philosophical-musings): Terlibat dalam diskusi filosofis mendalam dan eksperimen pemikiran.
- [Pembersih PII](https://docs.anthropic.com/id/prompt-library/pii-purifier): Secara otomatis mendeteksi dan menghapus informasi yang dapat mengidentifikasi pribadi (PII) dari dokumen teks.
- [Kekuatan poliglot](https://docs.anthropic.com/id/prompt-library/polyglot-superpowers): Menerjemahkan teks dari bahasa apa pun ke bahasa apa pun.
- [Penyair Portmanteau](https://docs.anthropic.com/id/prompt-library/portmanteau-poet): Menggabungkan dua kata untuk menciptakan portmanteau baru yang bermakna.
- [Pro penamaan produk](https://docs.anthropic.com/id/prompt-library/product-naming-pro): Membuat nama produk yang menarik dari deskripsi dan kata kunci.
- [Pemoles tulisan](https://docs.anthropic.com/id/prompt-library/prose-polisher): Menyempurnakan dan meningkatkan konten tertulis dengan teknik dan saran penyuntingan tingkat lanjut.
- [Pun-dit](https://docs.anthropic.com/id/prompt-library/pun-dit): Menghasilkan permainan kata dan lelucon cerdas berdasarkan topik apapun.
- [Pembasmi bug Python](https://docs.anthropic.com/id/prompt-library/python-bug-buster): Mendeteksi dan memperbaiki bug dalam kode Python.
- [Pengklasifikasi ulasan](https://docs.anthropic.com/id/prompt-library/review-classifier): Mengkategorikan umpan balik ke dalam tag dan kategorisasi yang telah ditentukan sebelumnya.
- [Teka-teki untukku](https://docs.anthropic.com/id/prompt-library/riddle-me-this): Membuat teka-teki dan memandu pengguna menuju solusinya.
- [Simulator skenario fiksi ilmiah](https://docs.anthropic.com/id/prompt-library/sci-fi-scenario-simulator): Diskusikan dengan pengguna berbagai skenario fiksi ilmiah dan tantangan serta pertimbangan terkait.
- [Penyederhana tingkat dua](https://docs.anthropic.com/id/prompt-library/second-grade-simplifier): Membuat teks kompleks mudah dipahami untuk pelajar muda
- [Ahli Simile](https://docs.anthropic.com/id/prompt-library/simile-savant): Menghasilkan simile dari deskripsi dasar.
- [Orang bijak Sokratik](https://docs.anthropic.com/id/prompt-library/socratic-sage): Terlibat dalam percakapan gaya Sokratik mengenai topik yang diberikan pengguna.
- [Penyihir spreadsheet](https://docs.anthropic.com/id/prompt-library/spreadsheet-sorcerer): Menghasilkan spreadsheet CSV dengan berbagai jenis data.
- [Penyihir SQL](https://docs.anthropic.com/id/prompt-library/sql-sorcerer): Mengubah bahasa sehari-hari menjadi kueri SQL.
- [Teman bercerita](https://docs.anthropic.com/id/prompt-library/storytelling-sidekick): Berkolaborasi menciptakan cerita menarik dengan pengguna, menawarkan plot twist dan pengembangan karakter.
- [Konsultan perjalanan waktu](https://docs.anthropic.com/id/prompt-library/time-travel-consultant): Membantu pengguna memahami skenario perjalanan waktu hipotetis dan implikasinya.
- [Permainan Lidah](https://docs.anthropic.com/id/prompt-library/tongue-twister): Membuat permainan lidah yang menantang.
- [Generator trivia](https://docs.anthropic.com/id/prompt-library/trivia-generator): Menghasilkan pertanyaan trivia tentang berbagai topik dan memberikan petunjuk saat diperlukan.
- [Detektor nada tweet](https://docs.anthropic.com/id/prompt-library/tweet-tone-detector): Mendeteksi nada dan sentimen di balik tweet.
- [Inovator kebugaran VR](https://docs.anthropic.com/id/prompt-library/vr-fitness-innovator): Brainstorming ide-ide kreatif untuk permainan kebugaran realitas virtual.
- [Pembuat situs web](https://docs.anthropic.com/id/prompt-library/website-wizard): Membuat situs web satu halaman berdasarkan spesifikasi pengguna.
- [API](https://docs.anthropic.com/id/release-notes/api): Ikuti pembaruan di seluruh API Anthropic dan Konsol Pengembang.
- [Aplikasi Claude](https://docs.anthropic.com/id/release-notes/claude-apps): Ikuti pembaruan di seluruh aplikasi Claude dari Anthropic.
- [Ikhtisar](https://docs.anthropic.com/id/release-notes/overview): Ikuti pembaruan di seluruh produk dan layanan Anthropic.
- [Prompt Sistem](https://docs.anthropic.com/id/release-notes/system-prompts): Lihat pembaruan pada prompt sistem inti di [Claude.ai](https://www.claude.ai) dan aplikasi Claude [iOS](http://anthropic.com/ios) dan [Android](http://anthropic.com/android).
- [Annulla un Batch di Messaggi (beta)](https://docs.anthropic.com/it/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [API Amazon Bedrock](https://docs.anthropic.com/it/api/claude-on-amazon-bedrock): I modelli Claude di Anthropic sono ora generalmente disponibili attraverso Amazon Bedrock.
- [API di Vertex AI](https://docs.anthropic.com/it/api/claude-on-vertex-ai): I modelli Claude di Anthropic sono ora generalmente disponibili attraverso [Vertex AI](https://cloud.google.com/vertex-ai).
- [SDK Client](https://docs.anthropic.com/it/api/client-sdks): Forniamo librerie in Python e TypeScript che rendono più semplice lavorare con l'API di Anthropic.
- [Create a Text Completion](https://docs.anthropic.com/it/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Crea un Batch di Messaggi (beta)](https://docs.anthropic.com/it/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Errori](https://docs.anthropic.com/it/api/errors)
- [Ottenere aiuto](https://docs.anthropic.com/it/api/getting-help): Abbiamo cercato di fornire le risposte alle domande più comuni in questa documentazione. Tuttavia, se hai bisogno di ulteriore supporto tecnico per l'utilizzo di Claude, l'API di Anthropic o qualsiasi dei nostri prodotti, puoi contattare il nostro team di supporto su [support.anthropic.com](https://support.anthropic.com).
- [Per iniziare](https://docs.anthropic.com/it/api/getting-started)
- [Indirizzi IP](https://docs.anthropic.com/it/api/ip-addresses): I servizi Anthropic sono disponibili su un intervallo fisso di indirizzi IP. Puoi aggiungere questi al tuo firewall per aprire la minima superficie necessaria per il traffico in uscita quando accedi all'API e alla Console Anthropic. Questi intervalli non cambieranno senza preavviso.
- [Elenco dei Lotti di Messaggi (beta)](https://docs.anthropic.com/it/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/it/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Esempi di Batch di Messaggi](https://docs.anthropic.com/it/api/messages-batch-examples): Esempi di utilizzo dell'API Message Batches
- [Conteggio dei token dei messaggi (beta)](https://docs.anthropic.com/it/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Esempi di messaggi](https://docs.anthropic.com/it/api/messages-examples): Esempi di richieste e risposte per l'API Messages
- [Messaggi in Streaming](https://docs.anthropic.com/it/api/messages-streaming)
- [Migrazione dalle Text Completions](https://docs.anthropic.com/it/api/migrating-from-text-completions-to-messages): Migrazione dalle Text Completions ai Messages
- [Validazione del prompt](https://docs.anthropic.com/it/api/prompt-validation): Con Text Completions
- [Limiti di utilizzo](https://docs.anthropic.com/it/api/rate-limits): Per mitigare l'uso improprio e gestire la capacità della nostra API, abbiamo implementato limiti su quanto un'organizzazione può utilizzare l'API di Claude.
- [Recupera i Risultati del Batch di Messaggi (beta)](https://docs.anthropic.com/it/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Recupera un Batch di Messaggi (beta)](https://docs.anthropic.com/it/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Completamenti di Testo in Streaming](https://docs.anthropic.com/it/api/streaming)
- [Regioni supportate](https://docs.anthropic.com/it/api/supported-regions): Ecco i paesi, le regioni e i territori da cui possiamo attualmente supportare l'accesso:
- [Versioni](https://docs.anthropic.com/it/api/versioning): Quando si effettuano richieste API, è necessario inviare un'intestazione di richiesta `anthropic-version`. Per esempio, `anthropic-version: 2023-06-01`. Se stai utilizzando le nostre [librerie client](/it/api/client-libraries), questo viene gestito automaticamente per te.
- [Aggiornamenti Luglio 2024](https://docs.anthropic.com/it/developer-newsletter/july2024): 28 Luglio 2024
- [Panoramica](https://docs.anthropic.com/it/developer-newsletter/overview): Esplora aggiornamenti mensili, approfondimenti tecnici, best practice e storie di successo per migliorare le tue integrazioni con Claude
- [Modelli](https://docs.anthropic.com/it/docs/about-claude/models): Claude è una famiglia di modelli linguistici di ultima generazione sviluppati da Anthropic. Questa guida presenta i nostri modelli e confronta le loro prestazioni con i modelli precedenti.
- [Sicurezza e conformità](https://docs.anthropic.com/it/docs/about-claude/security-compliance)
- [Moderazione dei contenuti](https://docs.anthropic.com/it/docs/about-claude/use-case-guides/content-moderation): La moderazione dei contenuti è un aspetto fondamentale per mantenere un ambiente sicuro, rispettoso e produttivo nelle applicazioni digitali. In questa guida, discuteremo come Claude può essere utilizzato per moderare i contenuti all'interno della tua applicazione digitale.
- [Agente di assistenza clienti](https://docs.anthropic.com/it/docs/about-claude/use-case-guides/customer-support-chat): Questa guida spiega come sfruttare le avanzate capacità conversazionali di Claude per gestire le richieste dei clienti in tempo reale, fornendo supporto 24/7, riducendo i tempi di attesa e gestendo elevati volumi di supporto con risposte accurate e interazioni positive.
- [Riassunto legale](https://docs.anthropic.com/it/docs/about-claude/use-case-guides/legal-summarization): Questa guida illustra come sfruttare le capacità avanzate di elaborazione del linguaggio naturale di Claude per riassumere in modo efficiente i documenti legali, estraendo le informazioni chiave e accelerando la ricerca legale. Con Claude, puoi semplificare la revisione dei contratti, la preparazione del contenzioso e il lavoro normativo, risparmiando tempo e garantendo precisione nei tuoi processi legali.
- [Guide per i casi d'uso comuni](https://docs.anthropic.com/it/docs/about-claude/use-case-guides/overview)
- [Instradamento dei ticket](https://docs.anthropic.com/it/docs/about-claude/use-case-guides/ticket-routing): Questa guida spiega come sfruttare le capacità avanzate di comprensione del linguaggio naturale di Claude per classificare i ticket di supporto clienti su larga scala in base all'intento del cliente, all'urgenza, alla prioritizzazione, al profilo del cliente e altro ancora.
- [Add-on per Google Sheets](https://docs.anthropic.com/it/docs/build-with-claude/claude-for-sheets): L'[estensione Claude per Sheets](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) integra Claude in Google Sheets, permettendo di eseguire interazioni con Claude direttamente nelle celle.
- [Uso del computer (beta)](https://docs.anthropic.com/it/docs/build-with-claude/computer-use): Il modello Claude 3.5 Sonnet aggiornato è in grado di interagire con [strumenti](/it/docs/build-with-claude/tool-use) che possono manipolare un ambiente desktop del computer.
- [Definisci i tuoi criteri di successo](https://docs.anthropic.com/it/docs/build-with-claude/define-success)
- [Creare valutazioni empiriche solide](https://docs.anthropic.com/it/docs/build-with-claude/develop-tests)
- [Embeddings](https://docs.anthropic.com/it/docs/build-with-claude/embeddings): Gli embeddings di testo sono rappresentazioni numeriche del testo che consentono di misurare la similarità semantica. Questa guida introduce gli embeddings, le loro applicazioni e come utilizzare i modelli di embedding per attività come ricerca, raccomandazioni e rilevamento di anomalie.
- [Batch di Messaggi (beta)](https://docs.anthropic.com/it/docs/build-with-claude/message-batches)
- [Supporto PDF (beta)](https://docs.anthropic.com/it/docs/build-with-claude/pdf-support): I modelli Claude 3.5 Sonnet ora supportano l'input PDF e comprendono sia il testo che il contenuto visivo all'interno dei documenti.
- [Cache delle prompt (beta)](https://docs.anthropic.com/it/docs/build-with-claude/prompt-caching)
- [Sii chiaro, diretto e dettagliato](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Lascia pensare Claude (prompt con catena di pensiero) per aumentare le prestazioni](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Concatena prompt complessi per prestazioni migliori](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Suggerimenti per il prompt con contesto lungo](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Usa esempi (prompt multishot) per guidare il comportamento di Claude](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Panoramica dell'ingegneria dei prompt](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/overview)
- [Precompila la risposta di Claude per un maggiore controllo dell'output](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Genera automaticamente modelli di prompt di prima bozza](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Usa il nostro strumento di miglioramento dei prompt per ottimizzare i tuoi prompt](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Utilizzare template di prompt e variabili](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Assegnare un ruolo a Claude con un prompt di sistema](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/system-prompts)
- [Usa i tag XML per strutturare i tuoi prompt](https://docs.anthropic.com/it/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Generazione di testo](https://docs.anthropic.com/it/docs/build-with-claude/text-generation): Claude eccelle in un'ampia gamma di attività basate sul testo. Claude è stato addestrato per elaborare codice, prosa e altri input in linguaggio naturale, e fornire output testuali in risposta.
- [Conteggio dei token (beta)](https://docs.anthropic.com/it/docs/build-with-claude/token-counting)
- [Utilizzo degli strumenti (chiamate di funzione)](https://docs.anthropic.com/it/docs/build-with-claude/tool-use)
- [Visione](https://docs.anthropic.com/it/docs/build-with-claude/vision): La famiglia di modelli Claude 3 è dotata di nuove capacità di visione che permettono a Claude di comprendere e analizzare le immagini, aprendo possibilità entusiasmanti per l'interazione multimodale.
- [Configurazione iniziale](https://docs.anthropic.com/it/docs/initial-setup): Impariamo come utilizzare l'API di Anthropic per costruire con Claude.
- [Introduzione a Claude](https://docs.anthropic.com/it/docs/intro-to-claude): Claude è una famiglia di [modelli di IA altamente performanti e intelligenti](/it/docs/about-claude/models) creati da Anthropic. Mentre Claude è potente ed estensibile, è anche l'IA più affidabile e attendibile disponibile. Segue protocolli critici, commette meno errori ed è resistente ai tentativi di elusione—permettendo ai [clienti enterprise](https://www.anthropic.com/customers) di costruire le applicazioni più sicure basate sull'IA su larga scala.
- [Informativa sulla Privacy di Anthropic](https://docs.anthropic.com/it/docs/legal-center/privacy)
- [Ricettario Anthropic](https://docs.anthropic.com/it/docs/resources/cookbook)
- [Corsi Anthropic](https://docs.anthropic.com/it/docs/resources/courses)
- [Glossario](https://docs.anthropic.com/it/docs/resources/glossary): Questi concetti non sono esclusivi dei modelli linguistici di Anthropic, ma presentiamo un breve riepilogo dei termini chiave qui sotto.
- [Scheda modello Claude 3](https://docs.anthropic.com/it/docs/resources/model-card)
- [Dismissione dei Modelli](https://docs.anthropic.com/it/docs/resources/model-deprecations)
- [Stato del sistema](https://docs.anthropic.com/it/docs/resources/status)
- [Utilizzo dello Strumento di Valutazione](https://docs.anthropic.com/it/docs/test-and-evaluate/eval-tool): La [Anthropic Console](https://console.anthropic.com/dashboard) include uno **Strumento di valutazione** che ti permette di testare i tuoi prompt in vari scenari.
- [Aumentare la coerenza dell'output (modalità JSON)](https://docs.anthropic.com/it/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Mantenere Claude nel personaggio con il role prompting e il prefilling](https://docs.anthropic.com/it/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Mitigare i jailbreak e le iniezioni di prompt](https://docs.anthropic.com/it/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Ridurre le allucinazioni](https://docs.anthropic.com/it/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Riduzione della latenza](https://docs.anthropic.com/it/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Ridurre la fuga di prompt](https://docs.anthropic.com/it/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Benvenuti a Claude](https://docs.anthropic.com/it/docs/welcome): Claude è una piattaforma di intelligenza artificiale altamente performante, affidabile e intelligente creata da Anthropic. Claude eccelle in compiti che coinvolgono linguaggio, ragionamento, analisi, programmazione e altro ancora.
- [null](https://docs.anthropic.com/it/home)
- [Editor adattivo](https://docs.anthropic.com/it/prompt-library/adaptive-editor): Riscrivere il testo seguendo le istruzioni dell'utente, come con un tono, un pubblico o uno stile diverso.
- [Analista di codici aeroportuali](https://docs.anthropic.com/it/prompt-library/airport-code-analyst): Trova ed estrai codici aeroportuali dal testo.
- [Antropologo alieno](https://docs.anthropic.com/it/prompt-library/alien-anthropologist): Analizza la cultura e i costumi umani dalla prospettiva di un antropologo alieno
- [Alchimista dell'allitterazione](https://docs.anthropic.com/it/prompt-library/alliteration-alchemist): Genera frasi e proposizioni allitterative per qualsiasi argomento dato.
- [Le trasmissioni di Babel](https://docs.anthropic.com/it/prompt-library/babels-broadcasts): Crea tweet di annunci di prodotti convincenti nelle 10 lingue più parlate al mondo
- [Costruttore di brand](https://docs.anthropic.com/it/prompt-library/brand-builder): Crea un brief di design per un'identità di marca olistica.
- [Coach di carriera](https://docs.anthropic.com/it/prompt-library/career-coach): Partecipa a conversazioni di role-play con un coach di carriera AI.
- [Cita le tue fonti](https://docs.anthropic.com/it/prompt-library/cite-your-sources): Ottieni risposte alle domande sul contenuto di un documento con citazioni pertinenti a supporto della risposta.
- [Chiarificatore di codice](https://docs.anthropic.com/it/prompt-library/code-clarifier): Semplifica e spiega il codice complesso in un linguaggio semplice
- [Consulente di codice](https://docs.anthropic.com/it/prompt-library/code-consultant): Suggerisce miglioramenti per ottimizzare le prestazioni del codice Python.
- [Chiaroveggente aziendale](https://docs.anthropic.com/it/prompt-library/corporate-clairvoyant): Estrai approfondimenti, identifica rischi e distilla informazioni chiave da lunghi report aziendali in un unico memo.
- [Battute Cosmiche](https://docs.anthropic.com/it/prompt-library/cosmic-keystrokes): Genera un gioco interattivo di digitazione veloce in un singolo file HTML, con gameplay a scorrimento laterale e stile Tailwind CSS.
- [Convertitore CSV](https://docs.anthropic.com/it/prompt-library/csv-converter): Converti dati da vari formati (JSON, XML, ecc.) in file CSV correttamente formattati.
- [Creatore culinario](https://docs.anthropic.com/it/prompt-library/culinary-creator): Suggerisci idee per ricette in base agli ingredienti disponibili e alle preferenze alimentari dell'utente.
- [Organizzatore di dati](https://docs.anthropic.com/it/prompt-library/data-organizer): Trasforma testo non strutturato in tabelle JSON personalizzate.
- [Decodificatore di direzioni](https://docs.anthropic.com/it/prompt-library/direction-decoder): Trasforma il linguaggio naturale in istruzioni passo dopo passo.
- [Interprete dei sogni](https://docs.anthropic.com/it/prompt-library/dream-interpreter): Offre interpretazioni e approfondimenti sul simbolismo dei sogni dell'utente
- [Stimatore di efficienza](https://docs.anthropic.com/it/prompt-library/efficiency-estimator): Calcola la complessità temporale di funzioni e algoritmi.
- [Estrattore di email](https://docs.anthropic.com/it/prompt-library/email-extractor): Estrai gli indirizzi email da un documento in un elenco formattato JSON.
- [Codificatore emoji](https://docs.anthropic.com/it/prompt-library/emoji-encoder): Converti testo semplice in messaggi divertenti ed espressivi con emoji.
- [Navigatore di dilemmi etici](https://docs.anthropic.com/it/prompt-library/ethical-dilemma-navigator): Aiuta l'utente a riflettere su complessi dilemmi etici e fornisce diverse prospettive
- [Esperto di formule Excel](https://docs.anthropic.com/it/prompt-library/excel-formula-expert): Crea formule Excel basate su calcoli o manipolazioni di dati descritti dall'utente.
- [Fabbricatore di funzioni](https://docs.anthropic.com/it/prompt-library/function-fabricator): Crea funzioni Python basate su specifiche dettagliate.
- [Consulente di moda futuristica](https://docs.anthropic.com/it/prompt-library/futuristic-fashion-advisor): Suggerisce tendenze e stili di moda all'avanguardia in base alle preferenze specifiche dell'utente
- [Diventa bravo con Git](https://docs.anthropic.com/it/prompt-library/git-gud): Genera comandi Git appropriati basati su azioni di controllo versione descritte dall'utente.
- [Programmatore di script per Google Apps](https://docs.anthropic.com/it/prompt-library/google-apps-scripter): Genera script di Google Apps per completare attività in base ai requisiti dell'utente
- [Guru della valutazione](https://docs.anthropic.com/it/prompt-library/grading-guru): Confronta e valuta la qualità dei testi scritti in base a criteri e standard definiti dall'utente.
- [Genio della grammatica](https://docs.anthropic.com/it/prompt-library/grammar-genie): Trasforma frasi grammaticalmente scorrette in inglese corretto.
- [Hal l'assistente umorista](https://docs.anthropic.com/it/prompt-library/hal-the-humorous-helper): Chatta con un'IA esperta che ha un lato sarcastico
- [Illuminatore di modi di dire](https://docs.anthropic.com/it/prompt-library/idiom-illuminator): Spiega il significato e l'origine di modi di dire e proverbi comuni
- [Generatore di domande per colloqui](https://docs.anthropic.com/it/prompt-library/interview-question-crafter): Genera domande per i colloqui.
- [Legenda LaTeX](https://docs.anthropic.com/it/prompt-library/latex-legend): Scrivi documenti LaTeX, generando codice per equazioni matematiche, tabelle e altro.
- [Pianificatore di lezioni](https://docs.anthropic.com/it/prompt-library/lesson-planner): Crea piani di lezione approfonditi su qualsiasi materia
- [Libreria](https://docs.anthropic.com/it/prompt-library/library)
- [Moderatore esperto](https://docs.anthropic.com/it/prompt-library/master-moderator): Valuta gli input degli utenti per potenziali contenuti dannosi o illegali
- [Scrivano di riunioni](https://docs.anthropic.com/it/prompt-library/meeting-scribe): Distilla le riunioni in sintesi concise che includono argomenti di discussione, punti chiave e azioni da intraprendere.
- [Memo maestro](https://docs.anthropic.com/it/prompt-library/memo-maestro): Componi memo aziendali completi basati su punti chiave
- [Mentore di mindfulness](https://docs.anthropic.com/it/prompt-library/mindfulness-mentor): Guida l'utente attraverso esercizi e tecniche di mindfulness per la riduzione dello stress.
- [Coloratore di stati d'animo](https://docs.anthropic.com/it/prompt-library/mood-colorizer): Trasforma le descrizioni testuali degli stati d'animo in codici HEX corrispondenti.
- [Musa motivazionale](https://docs.anthropic.com/it/prompt-library/motivational-muse): Fornisce messaggi motivazionali personalizzati e affermazioni basate sull'input dell'utente
- [Creatore di neologismi](https://docs.anthropic.com/it/prompt-library/neologism-creator): Inventa nuove parole e fornisce le loro definizioni basate su concetti o idee fornite dall'utente
- [Valutatore di prospettive](https://docs.anthropic.com/it/prompt-library/perspectives-ponderer): Valuta i pro e i contro di un argomento fornito dall'utente
- [Riflessioni filosofiche](https://docs.anthropic.com/it/prompt-library/philosophical-musings): Impegnati in profonde discussioni filosofiche ed esperimenti mentali
- [Purificatore PII](https://docs.anthropic.com/it/prompt-library/pii-purifier): Rileva e rimuove automaticamente le informazioni personali identificabili (PII) dai documenti di testo.
- [Superpoteri poliglotti](https://docs.anthropic.com/it/prompt-library/polyglot-superpowers): Traduci testo da qualsiasi lingua in qualsiasi lingua.
- [Poeta dei portmanteau](https://docs.anthropic.com/it/prompt-library/portmanteau-poet): Unisci due parole per creare un nuovo portmanteau significativo.
- [Professionista dei nomi di prodotti](https://docs.anthropic.com/it/prompt-library/product-naming-pro): Crea nomi di prodotti accattivanti da descrizioni e parole chiave
- [Perfezionatore di prosa](https://docs.anthropic.com/it/prompt-library/prose-polisher): Raffina e migliora i contenuti scritti con tecniche e suggerimenti avanzati di copyediting.
- [Pun-dit](https://docs.anthropic.com/it/prompt-library/pun-dit): Genera giochi di parole e battute intelligenti su qualsiasi argomento.
- [Correttore di bug Python](https://docs.anthropic.com/it/prompt-library/python-bug-buster): Rileva e correggi i bug nel codice Python.
- [Classificatore di recensioni](https://docs.anthropic.com/it/prompt-library/review-classifier): Categorizza i feedback in tag e categorizzazioni predefinite.
- [Indovinami questo](https://docs.anthropic.com/it/prompt-library/riddle-me-this): Genera indovinelli e guida l'utente verso le soluzioni.
- [Simulatore di scenari di fantascienza](https://docs.anthropic.com/it/prompt-library/sci-fi-scenario-simulator): Discuti con l'utente vari scenari di fantascienza e le relative sfide e considerazioni.
- [Semplificatore per la seconda elementare](https://docs.anthropic.com/it/prompt-library/second-grade-simplifier): Rendi i testi complessi facili da capire per i giovani studenti
- [Maestro delle similitudini](https://docs.anthropic.com/it/prompt-library/simile-savant): Genera similitudini da descrizioni di base
- [Saggio socratico](https://docs.anthropic.com/it/prompt-library/socratic-sage): Impegnati in una conversazione in stile socratico su un argomento fornito dall'utente
- [Stregone dei fogli di calcolo](https://docs.anthropic.com/it/prompt-library/spreadsheet-sorcerer): Genera fogli di calcolo CSV con vari tipi di dati.
- [Stregone SQL](https://docs.anthropic.com/it/prompt-library/sql-sorcerer): Trasforma il linguaggio quotidiano in query SQL.
- [Compagno di narrazione](https://docs.anthropic.com/it/prompt-library/storytelling-sidekick): Crea storie coinvolgenti in collaborazione con l'utente, offrendo colpi di scena e sviluppo dei personaggi.
- [Consulente di viaggi nel tempo](https://docs.anthropic.com/it/prompt-library/time-travel-consultant): Aiuta l'utente a navigare scenari ipotetici di viaggi nel tempo e le loro implicazioni.
- [Scioglilingua](https://docs.anthropic.com/it/prompt-library/tongue-twister): Crea scioglilingua impegnativi.
- [Generatore di quiz](https://docs.anthropic.com/it/prompt-library/trivia-generator): Genera domande di quiz su un'ampia gamma di argomenti e fornisce suggerimenti quando necessario.
- [Rilevatore del tono dei tweet](https://docs.anthropic.com/it/prompt-library/tweet-tone-detector): Rileva il tono e il sentimento dietro i tweet.
- [Innovatore del fitness in VR](https://docs.anthropic.com/it/prompt-library/vr-fitness-innovator): Sviluppare idee creative per giochi di fitness in realtà virtuale.
- [Creatore di siti web](https://docs.anthropic.com/it/prompt-library/website-wizard): Crea siti web di una pagina in base alle specifiche dell'utente.
- [API](https://docs.anthropic.com/it/release-notes/api): Segui gli aggiornamenti dell'API di Anthropic e della Console per Sviluppatori.
- [App di Claude](https://docs.anthropic.com/it/release-notes/claude-apps): Segui gli aggiornamenti delle applicazioni Claude di Anthropic.
- [Panoramica](https://docs.anthropic.com/it/release-notes/overview): Segui gli aggiornamenti dei prodotti e servizi di Anthropic.
- [Prompt di Sistema](https://docs.anthropic.com/it/release-notes/system-prompts): Vedi gli aggiornamenti ai prompt di sistema principali su [Claude.ai](https://www.claude.ai) e le app Claude [iOS](http://anthropic.com/ios) e [Android](http://anthropic.com/android).
- [メッセージバッチのキャンセル（ベータ版）](https://docs.anthropic.com/ja/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/ja/api/claude-on-amazon-bedrock): AnthropicのClaudeモデルがAmazon Bedrockで一般提供されるようになりました。
- [Vertex AI API](https://docs.anthropic.com/ja/api/claude-on-vertex-ai): AnthropicのClaudeモデルが[Vertex AI](https://cloud.google.com/vertex-ai)で一般提供されるようになりました。
- [クライアントSDK](https://docs.anthropic.com/ja/api/client-sdks): AnthropicのAPIをより簡単に使用できるようにするPythonとTypeScriptのライブラリを提供しています。
- [Create a Text Completion](https://docs.anthropic.com/ja/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [メッセージバッチの作成（ベータ版）](https://docs.anthropic.com/ja/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [エラー](https://docs.anthropic.com/ja/api/errors)
- [ヘルプの入手](https://docs.anthropic.com/ja/api/getting-help): これらのドキュメントで最も一般的な質問への回答を提供するよう努めています。ただし、Claude、Anthropic API、または当社の製品の使用に関してさらなる技術サポートが必要な場合は、[support.anthropic.com](https://support.anthropic.com)で当社のサポートチームにお問い合わせいただけます。
- [はじめに](https://docs.anthropic.com/ja/api/getting-started)
- [IPアドレス](https://docs.anthropic.com/ja/api/ip-addresses): Anthropicのサービスは固定のIPアドレス範囲で提供されています。これらのアドレスをファイアウォールに追加することで、AnthropicのAPIとコンソールにアクセスする際のアウトバウンドトラフィックの表面積を最小限に抑えることができます。これらの範囲は事前の通知なしには変更されません。
- [メッセージバッチの一覧表示（ベータ版）](https://docs.anthropic.com/ja/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/ja/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [メッセージバッチの例](https://docs.anthropic.com/ja/api/messages-batch-examples): メッセージバッチAPIの使用例
- [メッセージトークンのカウント（ベータ版）](https://docs.anthropic.com/ja/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [メッセージの例](https://docs.anthropic.com/ja/api/messages-examples): メッセージAPIのリクエストとレスポンスの例
- [メッセージのストリーミング](https://docs.anthropic.com/ja/api/messages-streaming)
- [テキスト補完からの移行](https://docs.anthropic.com/ja/api/migrating-from-text-completions-to-messages): テキスト補完からメッセージへの移行
- [プロンプトの検証](https://docs.anthropic.com/ja/api/prompt-validation): テキスト補完に関して
- [利用制限](https://docs.anthropic.com/ja/api/rate-limits): APIの誤用を防ぎ、容量を管理するため、組織がClaude APIを使用できる量に制限を設けています。
- [メッセージバッチ結果の取得（ベータ版）](https://docs.anthropic.com/ja/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [メッセージバッチの取得（ベータ版）](https://docs.anthropic.com/ja/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [ストリーミングテキスト補完](https://docs.anthropic.com/ja/api/streaming)
- [対応地域](https://docs.anthropic.com/ja/api/supported-regions): 現在アクセスをサポートしている国、地域、領土は以下の通りです：
- [バージョン](https://docs.anthropic.com/ja/api/versioning): APIリクエストを行う際は、`anthropic-version`リクエストヘッダーを送信する必要があります。例えば、`anthropic-version: 2023-06-01`のようになります。[クライアントライブラリ](/ja/api/client-libraries)を使用している場合、これは自動的に処理されます。
- [2024年7月のアップデート](https://docs.anthropic.com/ja/developer-newsletter/july2024): 2024年7月28日
- [概要](https://docs.anthropic.com/ja/developer-newsletter/overview): 月次アップデート、エンジニアリングの詳細解説、ベストプラクティス、成功事例を通じてClaudeの統合を強化しましょう
- [モデル](https://docs.anthropic.com/ja/docs/about-claude/models): Claudeは、Anthropicによって開発された最先端の大規模言語モデルファミリーです。このガイドでは、当社のモデルを紹介し、レガシーモデルとの性能を比較します。
- [セキュリティとコンプライアンス](https://docs.anthropic.com/ja/docs/about-claude/security-compliance)
- [コンテンツモデレーション](https://docs.anthropic.com/ja/docs/about-claude/use-case-guides/content-moderation): コンテンツモデレーションは、デジタルアプリケーションにおいて安全で、敬意を払い、生産的な環境を維持するための重要な側面です。このガイドでは、デジタルアプリケーション内でコンテンツをモデレートするためにClaudeをどのように使用できるかについて説明します。
- [カスタマーサポートエージェント](https://docs.anthropic.com/ja/docs/about-claude/use-case-guides/customer-support-chat): このガイドでは、Claudeの高度な会話機能を活用してリアルタイムで顧客の問い合わせに対応し、24時間365日のサポートを提供し、待ち時間を短縮し、正確な回答とポジティブなやり取りで大量のサポート要求を管理する方法を説明します。
- [法的文書の要約](https://docs.anthropic.com/ja/docs/about-claude/use-case-guides/legal-summarization): このガイドでは、Claudeの高度な自然言語処理機能を活用して、法的文書を効率的に要約し、重要な情報を抽出し、法的調査を迅速化する方法を説明します。Claudeを使用することで、契約書のレビュー、訴訟の準備、規制関連の作業を効率化し、法的プロセスにおける時間を節約し、正確性を確保することができます。
- [一般的なユースケースのガイド](https://docs.anthropic.com/ja/docs/about-claude/use-case-guides/overview)
- [チケットルーティング](https://docs.anthropic.com/ja/docs/about-claude/use-case-guides/ticket-routing): このガイドでは、Claudeの高度な自然言語理解能力を活用して、顧客の意図、緊急性、優先順位、顧客プロファイルなどに基づいて、大規模なカスタマーサポートチケットを分類する方法について説明します。
- [Google Sheets アドオン](https://docs.anthropic.com/ja/docs/build-with-claude/claude-for-sheets): [Claude for Sheets 拡張機能](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257)は、Claudeを Google Sheetsに統合し、セル内で直接Claudeとのやり取りを実行できるようにします。
- [コンピューターの使用 (ベータ版)](https://docs.anthropic.com/ja/docs/build-with-claude/computer-use): アップグレードされたClaude 3.5 Sonnetモデルは、コンピューターデスクトップ環境を操作できる[ツール](/ja/docs/build-with-claude/tool-use)と対話することができます。
- [成功基準を定義する](https://docs.anthropic.com/ja/docs/build-with-claude/define-success)
- [強力な実証評価の作成](https://docs.anthropic.com/ja/docs/build-with-claude/develop-tests)
- [埋め込み](https://docs.anthropic.com/ja/docs/build-with-claude/embeddings): テキスト埋め込みは、意味的類似性を測定できるテキストの数値表現です。このガイドでは、埋め込み、その応用、および検索、レコメンデーション、異常検出などのタスクに埋め込みモデルを使用する方法を紹介します。
- [メッセージバッチ（ベータ版）](https://docs.anthropic.com/ja/docs/build-with-claude/message-batches)
- [プロンプトキャッシング (ベータ版)](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-caching)
- [明確で直接的で詳細な指示を出す](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Claudeに考えさせる（思考の連鎖プロンプト）でパフォーマンスを向上させる](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [より強力なパフォーマンスのための複雑なプロンプトのチェーン化](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/chain-prompts)
- [長文コンテキストのプロンプト作成のヒント](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/long-context-tips)
- [例（マルチショットプロンプト）を使用してClaudeの動作を導く](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [プロンプトエンジニアリングの概要](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/overview)
- [Claudeの応答を事前入力してより良い出力制御を実現する](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [最初のドラフトプロンプトテンプレートを自動生成する](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/prompt-generator)
- [プロンプトインプルーバーを使用してプロンプトを最適化する](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/prompt-improver)
- [プロンプトテンプレートと変数を使用する](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Claudeに役割を与えるシステムプロンプトの使用](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/system-prompts)
- [XMLタグを使用してプロンプトを構造化する](https://docs.anthropic.com/ja/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [テキスト生成](https://docs.anthropic.com/ja/docs/build-with-claude/text-generation): Claudeは幅広いテキストベースのタスクで優れた性能を発揮します。Claudeはコード、文章、その他の自然言語入力を受け取り、それに応じてテキスト出力を提供するように訓練されています。
- [トークンカウント（ベータ）](https://docs.anthropic.com/ja/docs/build-with-claude/token-counting)
- [ツール使用（関数呼び出し）](https://docs.anthropic.com/ja/docs/build-with-claude/tool-use)
- [ビジョン](https://docs.anthropic.com/ja/docs/build-with-claude/vision): Claude 3ファミリーのモデルには、Claudeが画像を理解し分析できる新しいビジョン機能が搭載されており、マルチモーダルな対話の可能性が広がります。
- [初期設定](https://docs.anthropic.com/ja/docs/initial-setup): ClaudeとAnthropicのAPIを使用する方法を学びましょう。
- [Claudeの紹介](https://docs.anthropic.com/ja/docs/intro-to-claude): Claudeは、Anthropicが開発した[高性能で知的なAIモデル群](/ja/docs/about-claude/models)です。Claudeは強力で拡張性がありながら、最も信頼性が高く確実なAIでもあります。重要なプロトコルに従い、ミスが少なく、ジェイルブレイクに対して耐性があるため、[企業のお客様](https://www.anthropic.com/customers)は最も安全なAI駆動のアプリケーションを大規模に構築することができます。
- [Anthropic プライバシーポリシー](https://docs.anthropic.com/ja/docs/legal-center/privacy)
- [Anthropic クックブック](https://docs.anthropic.com/ja/docs/resources/cookbook)
- [Anthropic コース](https://docs.anthropic.com/ja/docs/resources/courses)
- [用語集](https://docs.anthropic.com/ja/docs/resources/glossary): これらの概念はAnthropicの言語モデルに固有のものではありませんが、主要な用語の簡単な概要を以下に示します。
- [Claude 3モデルカード](https://docs.anthropic.com/ja/docs/resources/model-card)
- [モデルの廃止](https://docs.anthropic.com/ja/docs/resources/model-deprecations)
- [システムステータス](https://docs.anthropic.com/ja/docs/resources/status)
- [評価ツールの使用方法](https://docs.anthropic.com/ja/docs/test-and-evaluate/eval-tool): [Anthropic Console](https://console.anthropic.com/dashboard)には、様々なシナリオでプロンプトをテストできる**評価ツール**が搭載されています。
- [出力の一貫性を高める (JSONモード)](https://docs.anthropic.com/ja/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [ロールプロンプトとプリフィルでClaudeのキャラクターを維持する](https://docs.anthropic.com/ja/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [ジェイルブレイクとプロンプトインジェクションの緩和](https://docs.anthropic.com/ja/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [幻覚を減らす](https://docs.anthropic.com/ja/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [レイテンシーの削減](https://docs.anthropic.com/ja/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [プロンプトリークの削減](https://docs.anthropic.com/ja/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Claudeへようこそ](https://docs.anthropic.com/ja/docs/welcome): Claudeは、Anthropicが構築した高性能で信頼性が高く、インテリジェントなAIプラットフォームです。言語、推論、分析、コーディングなどのタスクに優れています。
- [null](https://docs.anthropic.com/ja/home)
- [アダプティブエディター](https://docs.anthropic.com/ja/prompt-library/adaptive-editor): ユーザーの指示に従って、異なるトーン、対象読者、スタイルなどでテキストを書き換えます。
- [空港コード分析ツール](https://docs.anthropic.com/ja/prompt-library/airport-code-analyst): テキストから空港コードを見つけて抽出します。
- [エイリアン人類学者](https://docs.anthropic.com/ja/prompt-library/alien-anthropologist): エイリアン人類学者の視点から人間の文化と習慣を分析する
- [頭韻の錬金術師](https://docs.anthropic.com/ja/prompt-library/alliteration-alchemist): 任意のテーマに対して頭韻を踏んだフレーズや文章を生成します。
- [バベルの放送](https://docs.anthropic.com/ja/prompt-library/babels-broadcasts): 世界で最も話されている10言語で魅力的な製品発表ツイートを作成します。
- [ブランドビルダー](https://docs.anthropic.com/ja/prompt-library/brand-builder): 包括的なブランドアイデンティティのためのデザインブリーフを作成する
- [キャリアコーチ](https://docs.anthropic.com/ja/prompt-library/career-coach): AIキャリアコーチとのロールプレイ会話を体験
- [出典を引用する](https://docs.anthropic.com/ja/prompt-library/cite-your-sources): 文書の内容に関する質問に対して、回答を裏付ける関連引用とともに回答を得ることができます。
- [コード解説ツール](https://docs.anthropic.com/ja/prompt-library/code-clarifier): 複雑なコードをわかりやすい言葉で説明します。
- [コードコンサルタント](https://docs.anthropic.com/ja/prompt-library/code-consultant): Pythonコードのパフォーマンスを最適化するための改善を提案します。
- [企業の洞察力](https://docs.anthropic.com/ja/prompt-library/corporate-clairvoyant): 長い企業報告書から洞察を抽出し、リスクを特定し、重要な情報を1つのメモにまとめます。
- [コズミックキーストローク](https://docs.anthropic.com/ja/prompt-library/cosmic-keystrokes): サイドスクロールのゲームプレイとTailwind CSSスタイリングを特徴とする、インタラクティブなタイピングスピードゲームを単一のHTMLファイルで生成します。
- [CSVコンバーター](https://docs.anthropic.com/ja/prompt-library/csv-converter): 様々なフォーマット（JSON、XMLなど）のデータを適切にフォーマットされたCSVファイルに変換します。
- [料理クリエーター](https://docs.anthropic.com/ja/prompt-library/culinary-creator): ユーザーの手持ちの食材と食事制限に基づいてレシピのアイデアを提案します。
- [データオーガナイザー](https://docs.anthropic.com/ja/prompt-library/data-organizer): 非構造化テキストをカスタマイズされたJSONテーブルに変換します。
- [方向性デコーダー](https://docs.anthropic.com/ja/prompt-library/direction-decoder): 自然言語を段階的な指示に変換します。
- [夢の解釈者](https://docs.anthropic.com/ja/prompt-library/dream-interpreter): ユーザーの夢の象徴性について解釈とインサイトを提供します
- [効率性評価ツール](https://docs.anthropic.com/ja/prompt-library/efficiency-estimator): 関数とアルゴリズムの時間計算量を計算します。
- [メールアドレス抽出ツール](https://docs.anthropic.com/ja/prompt-library/email-extractor): ドキュメントからメールアドレスをJSON形式のリストとして抽出します。
- [絵文字エンコーダー](https://docs.anthropic.com/ja/prompt-library/emoji-encoder): プレーンテキストを楽しく表現豊かな絵文字メッセージに変換します。
- [倫理的ジレンマナビゲーター](https://docs.anthropic.com/ja/prompt-library/ethical-dilemma-navigator): 複雑な倫理的ジレンマを考え抜き、異なる視点を提供するためのサポート
- [Excel数式エキスパート](https://docs.anthropic.com/ja/prompt-library/excel-formula-expert): ユーザーが説明する計算やデータ操作に基づいてExcel数式を作成します。
- [関数ファブリケーター](https://docs.anthropic.com/ja/prompt-library/function-fabricator): 詳細な仕様に基づいてPython関数を作成します。
- [未来的なファッションアドバイザー](https://docs.anthropic.com/ja/prompt-library/futuristic-fashion-advisor): ユーザーの具体的な好みに合わせて、アバンギャルドなファッショントレンドとスタイルを提案します。
- [Git上達](https://docs.anthropic.com/ja/prompt-library/git-gud): ユーザーが説明するバージョン管理アクションに基づいて適切なGitコマンドを生成します。
- [Google Apps スクリプター](https://docs.anthropic.com/ja/prompt-library/google-apps-scripter): ユーザーの要件に基づいてタスクを完了するための Google Apps スクリプトを生成します。
- [採点の達人](https://docs.anthropic.com/ja/prompt-library/grading-guru): ユーザー定義の基準と標準に基づいて、文章の質を比較・評価します。
- [文法の魔人](https://docs.anthropic.com/ja/prompt-library/grammar-genie): 文法的に不正確な文を適切な英語に変換します。
- [ユーモアのあるヘルパー、ハル](https://docs.anthropic.com/ja/prompt-library/hal-the-humorous-helper): 皮肉な一面を持つ知識豊富なAIとチャットしましょう
- [慣用句解説](https://docs.anthropic.com/ja/prompt-library/idiom-illuminator): 一般的な慣用句や諺の意味と由来を説明します。
- [面接質問作成ツール](https://docs.anthropic.com/ja/prompt-library/interview-question-crafter): 面接のための質問を生成します。
- [LaTeX 凡例](https://docs.anthropic.com/ja/prompt-library/latex-legend): 数式、表などのコードを生成する LaTeX 文書の作成方法
- [レッスンプランナー](https://docs.anthropic.com/ja/prompt-library/lesson-planner): あらゆる科目の詳細な授業計画を作成します。
- [ライブラリ](https://docs.anthropic.com/ja/prompt-library/library)
- [マスターモデレーター](https://docs.anthropic.com/ja/prompt-library/master-moderator): ユーザー入力に有害または違法なコンテンツが含まれていないか評価します。
- [会議の書記](https://docs.anthropic.com/ja/prompt-library/meeting-scribe): 議論のトピック、重要なポイント、アクションアイテムを含む簡潔な会議の要約を作成します。
- [メモの達人](https://docs.anthropic.com/ja/prompt-library/memo-maestro): 要点に基づいて包括的な社内メモを作成します。
- [マインドフルネスメンター](https://docs.anthropic.com/ja/prompt-library/mindfulness-mentor): ストレス軽減のためのマインドフルネスエクササイズとテクニックをユーザーに指導します
- [ムードカラーライザー](https://docs.anthropic.com/ja/prompt-library/mood-colorizer): ムードの文章表現を対応するHEXコードに変換します。
- [モチベーションミューズ](https://docs.anthropic.com/ja/prompt-library/motivational-muse): ユーザーの入力に基づいて、個別のモチベーションメッセージと肯定的なメッセージを提供します。
- [新語創造者](https://docs.anthropic.com/ja/prompt-library/neologism-creator): ユーザーが提供する概念やアイデアに基づいて新しい言葉を作り出し、その定義を提供します。
- [視点の考察者](https://docs.anthropic.com/ja/prompt-library/perspectives-ponderer): ユーザーが提供したトピックの長所と短所を検討します。
- [哲学的考察](https://docs.anthropic.com/ja/prompt-library/philosophical-musings): 深い哲学的議論と思考実験に取り組む
- [PII 除去ツール](https://docs.anthropic.com/ja/prompt-library/pii-purifier): テキスト文書から個人を特定できる情報（PII）を自動的に検出して削除します。
- [多言語スーパーパワー](https://docs.anthropic.com/ja/prompt-library/polyglot-superpowers): あらゆる言語からあらゆる言語へテキストを翻訳します。
- [ポートマントー詩人](https://docs.anthropic.com/ja/prompt-library/portmanteau-poet): 2つの単語を組み合わせて、新しい意味のあるポートマントーを作成します。
- [製品ネーミングのプロ](https://docs.anthropic.com/ja/prompt-library/product-naming-pro): 説明とキーワードから魅力的な製品名を作成します。
- [文章の洗練](https://docs.anthropic.com/ja/prompt-library/prose-polisher): 高度な校正技術と提案で文章を改善・洗練させます
- [パンディット](https://docs.anthropic.com/ja/prompt-library/pun-dit): 任意のトピックに基づいて巧みな言葉遊びとダジャレを生成します。
- [Pythonバグバスター](https://docs.anthropic.com/ja/prompt-library/python-bug-buster): Pythonコードのバグを検出して修正します。
- [レビュー分類器](https://docs.anthropic.com/ja/prompt-library/review-classifier): フィードバックを事前に指定されたタグとカテゴリーに分類します。
- [なぞなぞに答えてみよう](https://docs.anthropic.com/ja/prompt-library/riddle-me-this): なぞなぞを生成し、ユーザーを解答へと導きます。
- [SF シナリオシミュレーター](https://docs.anthropic.com/ja/prompt-library/sci-fi-scenario-simulator): ユーザーとさまざまなSFシナリオについて、関連する課題や考慮事項を議論します。
- [小学2年生向けの簡単化ツール](https://docs.anthropic.com/ja/prompt-library/second-grade-simplifier): 複雑なテキストを若い学習者が理解しやすいように変換します
- [直喩の達人](https://docs.anthropic.com/ja/prompt-library/simile-savant): 基本的な説明から直喩を生成します。
- [ソクラテス式の賢者](https://docs.anthropic.com/ja/prompt-library/socratic-sage): ユーザーが提供したトピックについてソクラテス式の対話を行います
- [スプレッドシートの魔術師](https://docs.anthropic.com/ja/prompt-library/spreadsheet-sorcerer): 様々な種類のデータを含むCSVスプレッドシートを生成します。
- [SQLソーサラー](https://docs.anthropic.com/ja/prompt-library/sql-sorcerer): 日常言語をSQLクエリに変換します。
- [ストーリーテリングのパートナー](https://docs.anthropic.com/ja/prompt-library/storytelling-sidekick): ユーザーと協力して魅力的なストーリーを作成し、プロットの展開やキャラクターの成長を提案します。
- [タイムトラベルコンサルタント](https://docs.anthropic.com/ja/prompt-library/time-travel-consultant): ユーザーが仮説的なタイムトラベルのシナリオとその影響について理解できるようサポートします。
- [早口言葉](https://docs.anthropic.com/ja/prompt-library/tongue-twister): 難しい早口言葉を作成します。
- [クイズジェネレーター](https://docs.anthropic.com/ja/prompt-library/trivia-generator): 幅広いトピックについてのクイズ問題を生成し、必要に応じてヒントを提供します。
- [ツイートのトーン検出器](https://docs.anthropic.com/ja/prompt-library/tweet-tone-detector): ツイートの背後にあるトーンと感情を検出します。
- [VRフィットネスイノベーター](https://docs.anthropic.com/ja/prompt-library/vr-fitness-innovator): バーチャルリアリティフィットネスゲームのための創造的なアイデアを考案する
- [ウェブサイトウィザード](https://docs.anthropic.com/ja/prompt-library/website-wizard): ユーザーの仕様に基づいてワンページウェブサイトを作成します。
- [API](https://docs.anthropic.com/ja/release-notes/api): AnthropicのAPIと開発者コンソールの更新情報をご確認ください。
- [Claude アプリ](https://docs.anthropic.com/ja/release-notes/claude-apps): Anthropicのクラウドアプリケーションの最新情報をお届けします。
- [概要](https://docs.anthropic.com/ja/release-notes/overview): Anthropicの製品とサービスの最新情報をご覧ください。
- [システムプロンプト](https://docs.anthropic.com/ja/release-notes/system-prompts): [Claude.ai](https://www.claude.ai)とClaude [iOS](http://anthropic.com/ios)および[Android](http://anthropic.com/android)アプリのコアシステムプロンプトの更新をご覧ください。
- [메시지 배치 취소하기 (베타)](https://docs.anthropic.com/ko/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/ko/api/claude-on-amazon-bedrock): Anthropic의 Claude 모델이 이제 Amazon Bedrock을 통해 일반적으로 사용 가능합니다.
- [Vertex AI API](https://docs.anthropic.com/ko/api/claude-on-vertex-ai): Anthropic의 Claude 모델이 이제 [Vertex AI](https://cloud.google.com/vertex-ai)를 통해 일반적으로 사용 가능합니다.
- [클라이언트 SDK](https://docs.anthropic.com/ko/api/client-sdks): Python과 TypeScript에서 Anthropic API를 더 쉽게 사용할 수 있게 해주는 라이브러리를 제공합니다.
- [Create a Text Completion](https://docs.anthropic.com/ko/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [메시지 배치 생성하기 (베타)](https://docs.anthropic.com/ko/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [오류](https://docs.anthropic.com/ko/api/errors)
- [도움 받기](https://docs.anthropic.com/ko/api/getting-help): 이 문서에서 가장 일반적인 질문에 대한 답변을 제공하고자 했습니다. 하지만 Claude, Anthropic API 또는 당사 제품 사용에 대한 추가 기술 지원이 필요한 경우 [support.anthropic.com](https://support.anthropic.com)에서 지원팀에 문의하실 수 있습니다.
- [시작하기](https://docs.anthropic.com/ko/api/getting-started)
- [IP 주소](https://docs.anthropic.com/ko/api/ip-addresses): Anthropic 서비스는 고정된 IP 주소 범위에서 운영됩니다. Anthropic API와 콘솔에 접근할 때 최소한의 송신 트래픽 영역을 열기 위해 이 주소들을 방화벽에 추가할 수 있습니다. 이 범위는 사전 공지 없이 변경되지 않습니다.
- [메시지 배치 목록 조회 (베타)](https://docs.anthropic.com/ko/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/ko/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [메시지 배치 예제](https://docs.anthropic.com/ko/api/messages-batch-examples): 메시지 배치 API 사용 예제
- [메시지 토큰 수 계산 (베타)](https://docs.anthropic.com/ko/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [메시지 예시](https://docs.anthropic.com/ko/api/messages-examples): 메시지 API의 요청 및 응답 예시
- [메시지 스트리밍](https://docs.anthropic.com/ko/api/messages-streaming)
- [텍스트 완성에서 마이그레이션하기](https://docs.anthropic.com/ko/api/migrating-from-text-completions-to-messages): 텍스트 완성에서 메시지로 마이그레이션하기
- [프롬프트 검증](https://docs.anthropic.com/ko/api/prompt-validation): 텍스트 완성 기능 관련
- [사용량 제한](https://docs.anthropic.com/ko/api/rate-limits): API의 오용을 방지하고 용량을 관리하기 위해 조직이 Claude API를 사용할 수 있는 한도를 설정했습니다.
- [메시지 배치 결과 검색 (베타)](https://docs.anthropic.com/ko/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [메시지 배치 검색 (베타)](https://docs.anthropic.com/ko/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [텍스트 완성 스트리밍](https://docs.anthropic.com/ko/api/streaming)
- [지원되는 지역](https://docs.anthropic.com/ko/api/supported-regions): 현재 접근이 가능한 국가, 지역 및 영토는 다음과 같습니다:
- [버전](https://docs.anthropic.com/ko/api/versioning): API 요청을 할 때는 `anthropic-version` 요청 헤더를 보내야 합니다. 예를 들어, `anthropic-version: 2023-06-01`과 같습니다. [클라이언트 라이브러리](/ko/api/client-libraries)를 사용하는 경우 이는 자동으로 처리됩니다.
- [2024년 7월 업데이트](https://docs.anthropic.com/ko/developer-newsletter/july2024): 2024년 7월 28일
- [개요](https://docs.anthropic.com/ko/developer-newsletter/overview): 월간 업데이트, 엔지니어링 심층 분석, 모범 사례 및 성공 사례를 통해 Claude 통합을 향상시키세요
- [모델](https://docs.anthropic.com/ko/docs/about-claude/models): Claude는 Anthropic이 개발한 최첨단 대규모 언어 모델군입니다. 이 가이드에서는 우리의 모델들을 소개하고 기존 모델들과의 성능을 비교합니다.
- [보안 및 규정 준수](https://docs.anthropic.com/ko/docs/about-claude/security-compliance)
- [콘텐츠 관리](https://docs.anthropic.com/ko/docs/about-claude/use-case-guides/content-moderation): 콘텐츠 관리는 디지털 애플리케이션에서 안전하고 존중되며 생산적인 환경을 유지하는 데 중요한 측면입니다. 이 가이드에서는 Claude를 사용하여 디지털 애플리케이션 내의 콘텐츠를 관리하는 방법을 설명합니다.
- [고객 지원 상담원](https://docs.anthropic.com/ko/docs/about-claude/use-case-guides/customer-support-chat): 이 가이드는 Claude의 고급 대화 기능을 활용하여 실시간으로 고객 문의를 처리하고, 24/7 지원을 제공하며, 대기 시간을 줄이고, 정확한 응답과 긍정적인 상호작용으로 대량의 지원 요청을 관리하는 방법을 설명합니다.
- [법률 문서 요약](https://docs.anthropic.com/ko/docs/about-claude/use-case-guides/legal-summarization): 이 가이드는 Claude의 고급 자연어 처리 기능을 활용하여 법률 문서를 효율적으로 요약하고, 핵심 정보를 추출하며, 법률 연구를 신속하게 처리하는 방법을 설명합니다. Claude를 사용하면 계약서 검토, 소송 준비 및 규제 업무를 간소화하여 시간을 절약하고 법적 절차의 정확성을 보장할 수 있습니다.
- [일반적인 사용 사례 가이드](https://docs.anthropic.com/ko/docs/about-claude/use-case-guides/overview)
- [티켓 라우팅](https://docs.anthropic.com/ko/docs/about-claude/use-case-guides/ticket-routing): 이 가이드는 고객 의도, 긴급성, 우선순위, 고객 프로필 등을 기반으로 대규모 고객 지원 티켓을 분류하기 위해 Claude의 고급 자연어 이해 기능을 활용하는 방법을 설명합니다.
- [Google Sheets 애드온](https://docs.anthropic.com/ko/docs/build-with-claude/claude-for-sheets): [Claude for Sheets 확장 프로그램](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257)은 Claude를 Google Sheets에 통합하여 셀에서 직접 Claude와 상호작용할 수 있게 해줍니다.
- [컴퓨터 사용 (베타)](https://docs.anthropic.com/ko/docs/build-with-claude/computer-use): 업그레이드된 Claude 3.5 Sonnet 모델은 데스크톱 환경을 조작할 수 있는 [도구](/ko/docs/build-with-claude/tool-use)와 상호 작용할 수 있습니다.
- [성공 기준 정의하기](https://docs.anthropic.com/ko/docs/build-with-claude/define-success)
- [강력한 실증적 평가 만들기](https://docs.anthropic.com/ko/docs/build-with-claude/develop-tests)
- [임베딩](https://docs.anthropic.com/ko/docs/build-with-claude/embeddings): 텍스트 임베딩은 의미적 유사성을 측정할 수 있게 해주는 텍스트의 수치적 표현입니다. 이 가이드는 임베딩, 그 응용 분야, 그리고 검색, 추천, 이상 탐지와 같은 작업을 위한 임베딩 모델 사용 방법을 소개합니다.
- [메시지 배치 (베타)](https://docs.anthropic.com/ko/docs/build-with-claude/message-batches)
- [PDF 지원 (베타)](https://docs.anthropic.com/ko/docs/build-with-claude/pdf-support): Claude 3.5 Sonnet 모델은 이제 PDF 입력을 지원하며 문서 내의 텍스트와 시각적 콘텐츠를 모두 이해합니다.
- [프롬프트 캐싱 (베타)](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-caching)
- [명확하고 직접적이며 상세하게](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Claude가 생각하도록 하여(사고 연쇄 프롬프팅) 성능을 향상시키기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [더 강력한 성능을 위한 복잡한 프롬프트 체이닝](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/chain-prompts)
- [긴 컨텍스트 프롬프팅 팁](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/long-context-tips)
- [예시(멀티샷 프롬프팅)를 사용하여 Claude의 동작 유도하기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [프롬프트 엔지니어링 개요](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/overview)
- [더 나은 출력 제어를 위한 Claude의 응답 미리 채우기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [자동으로 첫 초안 프롬프트 템플릿 생성하기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/prompt-generator)
- [프롬프트 개선기를 사용하여 프롬프트 최적화하기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/prompt-improver)
- [프롬프트 템플릿과 변수 사용하기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Claude에게 시스템 프롬프트로 역할 부여하기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/system-prompts)
- [XML 태그를 사용하여 프롬프트 구조화하기](https://docs.anthropic.com/ko/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [텍스트 생성](https://docs.anthropic.com/ko/docs/build-with-claude/text-generation): Claude는 광범위한 텍스트 기반 작업에서 뛰어난 성능을 보입니다. Claude는 코드, 산문 및 기타 자연어 입력을 받아들이고 이에 대한 텍스트 출력을 제공하도록 훈련되었습니다.
- [토큰 카운팅 (베타)](https://docs.anthropic.com/ko/docs/build-with-claude/token-counting)
- [도구 사용 (함수 호출)](https://docs.anthropic.com/ko/docs/build-with-claude/tool-use)
- [비전](https://docs.anthropic.com/ko/docs/build-with-claude/vision): Claude 3 제품군은 새로운 비전 기능을 통해 Claude가 이미지를 이해하고 분석할 수 있게 하여 멀티모달 상호작용을 위한 흥미로운 가능성을 열어줍니다.
- [초기 설정](https://docs.anthropic.com/ko/docs/initial-setup): Claude와 함께 구축하기 위해 Anthropic API를 사용하는 방법을 배워보겠습니다.
- [Claude 소개](https://docs.anthropic.com/ko/docs/intro-to-claude): Claude는 Anthropic이 개발한 [고성능 및 지능형 AI 모델](/ko/docs/about-claude/models) 제품군입니다. Claude는 강력하고 확장 가능하면서도 가장 신뢰할 수 있고 안정적인 AI입니다. 중요한 프로토콜을 따르고, 실수가 적으며, 해킹에 강한 내성을 가지고 있어 [기업 고객](https://www.anthropic.com/customers)이 대규모로 가장 안전한 AI 기반 애플리케이션을 구축할 수 있습니다.
- [Anthropic 개인정보 보호정책](https://docs.anthropic.com/ko/docs/legal-center/privacy)
- [Anthropic 요리책](https://docs.anthropic.com/ko/docs/resources/cookbook)
- [Anthropic 강좌](https://docs.anthropic.com/ko/docs/resources/courses)
- [용어집](https://docs.anthropic.com/ko/docs/resources/glossary): 이러한 개념들은 Anthropic의 언어 모델에만 국한된 것은 아니지만, 주요 용어들에 대한 간단한 요약을 제시합니다.
- [Claude 3 모델 카드](https://docs.anthropic.com/ko/docs/resources/model-card)
- [모델 폐기](https://docs.anthropic.com/ko/docs/resources/model-deprecations)
- [시스템 상태](https://docs.anthropic.com/ko/docs/resources/status)
- [평가 도구 사용하기](https://docs.anthropic.com/ko/docs/test-and-evaluate/eval-tool): [Anthropic Console](https://console.anthropic.com/dashboard)은 다양한 시나리오에서 프롬프트를 테스트할 수 있는 **평가 도구**를 제공합니다.
- [출력 일관성 높이기 (JSON 모드)](https://docs.anthropic.com/ko/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [역할 프롬프팅과 사전 채우기로 Claude의 캐릭터 유지하기](https://docs.anthropic.com/ko/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [탈옥 및 프롬프트 인젝션 완화](https://docs.anthropic.com/ko/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [환각 현상 줄이기](https://docs.anthropic.com/ko/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [지연 시간 감소](https://docs.anthropic.com/ko/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [프롬프트 유출 감소](https://docs.anthropic.com/ko/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Claude 시작하기](https://docs.anthropic.com/ko/docs/welcome): Claude는 Anthropic이 개발한 고성능의 신뢰할 수 있는 지능형 AI 플랫폼입니다. Claude는 언어, 추론, 분석, 코딩 등의 작업에서 탁월한 성능을 보입니다.
- [null](https://docs.anthropic.com/ko/home)
- [적응형 에디터](https://docs.anthropic.com/ko/prompt-library/adaptive-editor): 사용자가 제공한 지침에 따라 다른 톤, 청중, 또는 스타일로 텍스트를 다시 작성합니다.
- [공항 코드 분석기](https://docs.anthropic.com/ko/prompt-library/airport-code-analyst): 텍스트에서 공항 코드를 찾고 추출합니다.
- [외계인 인류학자](https://docs.anthropic.com/ko/prompt-library/alien-anthropologist): 외계인 인류학자의 관점에서 인간 문화와 관습을 분석합니다.
- [두운법 연금술사](https://docs.anthropic.com/ko/prompt-library/alliteration-alchemist): 주어진 주제에 대한 두운법 구문과 문장을 생성합니다.
- [바벨의 방송](https://docs.anthropic.com/ko/prompt-library/babels-broadcasts): 세계에서 가장 많이 사용되는 10개 언어로 매력적인 제품 발표 트윗을 작성하세요.
- [브랜드 빌더](https://docs.anthropic.com/ko/prompt-library/brand-builder): 전체적인 브랜드 아이덴티티를 위한 디자인 브리프를 작성하세요.
- [커리어 코치](https://docs.anthropic.com/ko/prompt-library/career-coach): AI 커리어 코치와 역할극 대화에 참여하세요.
- [출처 인용하기](https://docs.anthropic.com/ko/prompt-library/cite-your-sources): 문서 내용에 대한 질문에 대해 관련 인용을 포함한 답변을 받으세요.
- [코드 명확화 도구](https://docs.anthropic.com/ko/prompt-library/code-clarifier): 복잡한 코드를 쉬운 언어로 단순화하고 설명합니다.
- [코드 컨설턴트](https://docs.anthropic.com/ko/prompt-library/code-consultant): Python 코드 성능을 최적화하기 위한 개선사항을 제안합니다.
- [기업 통찰력](https://docs.anthropic.com/ko/prompt-library/corporate-clairvoyant): 긴 기업 보고서에서 통찰력을 추출하고, 위험을 식별하며, 핵심 정보를 단일 메모로 요약합니다.
- [우주의 키스트로크](https://docs.anthropic.com/ko/prompt-library/cosmic-keystrokes): 사이드 스크롤링 게임플레이와 Tailwind CSS 스타일링을 특징으로 하는 대화형 속도 타이핑 게임을 단일 HTML 파일로 생성합니다.
- [CSV 변환기](https://docs.anthropic.com/ko/prompt-library/csv-converter): 다양한 형식(JSON, XML 등)의 데이터를 올바르게 형식이 지정된 CSV 파일로 변환합니다.
- [요리 크리에이터](https://docs.anthropic.com/ko/prompt-library/culinary-creator): 사용자의 가용 재료와 식이 선호도를 기반으로 레시피 아이디어를 제안합니다.
- [데이터 정리기](https://docs.anthropic.com/ko/prompt-library/data-organizer): 비정형 텍스트를 맞춤형 JSON 테이블로 변환합니다.
- [방향 디코더](https://docs.anthropic.com/ko/prompt-library/direction-decoder): 자연어를 단계별 지침으로 변환합니다.
- [꿈 해석가](https://docs.anthropic.com/ko/prompt-library/dream-interpreter): 사용자의 꿈에 대한 상징성과 통찰력 있는 해석을 제공합니다.
- [효율성 추정기](https://docs.anthropic.com/ko/prompt-library/efficiency-estimator): 함수와 알고리즘의 시간 복잡도를 계산합니다.
- [이메일 추출기](https://docs.anthropic.com/ko/prompt-library/email-extractor): 문서에서 이메일 주소를 JSON 형식 목록으로 추출합니다.
- [이모지 인코더](https://docs.anthropic.com/ko/prompt-library/emoji-encoder): 일반 텍스트를 재미있고 표현력 있는 이모지 메시지로 변환합니다.
- [윤리적 딜레마 내비게이터](https://docs.anthropic.com/ko/prompt-library/ethical-dilemma-navigator): 사용자가 복잡한 윤리적 딜레마를 생각하고 다양한 관점을 제공하도록 돕습니다.
- [엑셀 수식 전문가](https://docs.anthropic.com/ko/prompt-library/excel-formula-expert): 사용자가 설명한 계산이나 데이터 조작을 기반으로 엑셀 수식을 생성합니다.
- [함수 제작기](https://docs.anthropic.com/ko/prompt-library/function-fabricator): 상세한 명세를 기반으로 Python 함수를 생성합니다.
- [미래지향적 패션 어드바이저](https://docs.anthropic.com/ko/prompt-library/futuristic-fashion-advisor): 사용자의 구체적인 선호도에 맞는 아방가르드 패션 트렌드와 스타일을 제안합니다.
- [Git 잘하기](https://docs.anthropic.com/ko/prompt-library/git-gud): 사용자가 설명한 버전 관리 작업을 기반으로 적절한 Git 명령어를 생성합니다.
- [Google 앱스 스크립터](https://docs.anthropic.com/ko/prompt-library/google-apps-scripter): 사용자 요구사항에 따라 작업을 완료하기 위한 Google 앱스 스크립트를 생성합니다.
- [채점 전문가](https://docs.anthropic.com/ko/prompt-library/grading-guru): 사용자가 정의한 기준과 표준에 따라 작성된 텍스트의 품질을 비교하고 평가합니다.
- [문법의 요정](https://docs.anthropic.com/ko/prompt-library/grammar-genie): 문법적으로 잘못된 문장을 올바른 영어로 변환합니다.
- [유머러스한 도우미 Hal](https://docs.anthropic.com/ko/prompt-library/hal-the-humorous-helper): 재치있는 면모를 가진 박식한 AI와 대화하세요.
- [관용구 해설가](https://docs.anthropic.com/ko/prompt-library/idiom-illuminator): 일반적인 관용구와 속담의 의미와 기원을 설명합니다.
- [면접 질문 생성기](https://docs.anthropic.com/ko/prompt-library/interview-question-crafter): 면접을 위한 질문을 생성합니다.
- [LaTeX 범례](https://docs.anthropic.com/ko/prompt-library/latex-legend): 수학 방정식, 표 등을 위한 코드를 생성하는 LaTeX 문서 작성하기
- [수업 계획 작성기](https://docs.anthropic.com/ko/prompt-library/lesson-planner): 모든 과목에 대한 심층적인 수업 계획을 작성하세요.
- [라이브러리](https://docs.anthropic.com/ko/prompt-library/library)
- [마스터 모더레이터](https://docs.anthropic.com/ko/prompt-library/master-moderator): 사용자 입력에서 잠재적으로 유해하거나 불법적인 콘텐츠를 평가합니다.
- [회의록 작성](https://docs.anthropic.com/ko/prompt-library/meeting-scribe): 회의 내용을 토론 주제, 주요 결과 및 실행 항목을 포함한 간결한 요약으로 정리합니다.
- [메모 마에스트로](https://docs.anthropic.com/ko/prompt-library/memo-maestro): 핵심 요점을 바탕으로 포괄적인 회사 메모를 작성합니다.
- [마음챙김 멘토](https://docs.anthropic.com/ko/prompt-library/mindfulness-mentor): 사용자를 마음챙김 운동과 스트레스 감소 기법으로 안내합니다.
- [감정 색상화](https://docs.anthropic.com/ko/prompt-library/mood-colorizer): 감정 설명을 해당하는 HEX 코드로 변환합니다.
- [동기부여 뮤즈](https://docs.anthropic.com/ko/prompt-library/motivational-muse): 사용자 입력을 기반으로 개인화된 동기부여 메시지와 긍정적인 확언을 제공합니다.
- [신조어 생성기](https://docs.anthropic.com/ko/prompt-library/neologism-creator): 사용자가 제공한 개념이나 아이디어를 바탕으로 새로운 단어를 만들고 그 정의를 제공합니다.
- [관점 분석기](https://docs.anthropic.com/ko/prompt-library/perspectives-ponderer): 사용자가 제공한 주제의 장단점을 분석합니다.
- [철학적 사색](https://docs.anthropic.com/ko/prompt-library/philosophical-musings): 깊이 있는 철학적 토론과 사고 실험에 참여하세요.
- [PII 제거기](https://docs.anthropic.com/ko/prompt-library/pii-purifier): 텍스트 문서에서 개인식별정보(PII)를 자동으로 감지하고 제거합니다.
- [다국어 초능력](https://docs.anthropic.com/ko/prompt-library/polyglot-superpowers): 모든 언어에서 모든 언어로 텍스트를 번역하세요.
- [포트만토 시인](https://docs.anthropic.com/ko/prompt-library/portmanteau-poet): 두 단어를 결합하여 새롭고 의미 있는 포트만토를 만듭니다.
- [제품 이름 전문가](https://docs.anthropic.com/ko/prompt-library/product-naming-pro): 설명과 키워드로부터 매력적인 제품 이름을 만듭니다.
- [문장 개선기](https://docs.anthropic.com/ko/prompt-library/prose-polisher): 고급 교정 기법과 제안을 통해 작성된 콘텐츠를 다듬고 개선합니다.
- [말장난꾼](https://docs.anthropic.com/ko/prompt-library/pun-dit): 주어진 주제에 대한 재치있는 말장난과 언어유희를 생성합니다.
- [Python 버그 버스터](https://docs.anthropic.com/ko/prompt-library/python-bug-buster): Python 코드의 버그를 감지하고 수정합니다.
- [리뷰 분류기](https://docs.anthropic.com/ko/prompt-library/review-classifier): 피드백을 사전 지정된 태그와 분류로 카테고리화합니다.
- [수수께끼를 풀어보세요](https://docs.anthropic.com/ko/prompt-library/riddle-me-this): 수수께끼를 생성하고 사용자가 해답을 찾도록 안내합니다
- [공상과학 시나리오 시뮬레이터](https://docs.anthropic.com/ko/prompt-library/sci-fi-scenario-simulator): 사용자와 함께 다양한 공상과학 시나리오와 관련된 도전과제 및 고려사항을 논의합니다.
- [초등학교 2학년 수준 단순화기](https://docs.anthropic.com/ko/prompt-library/second-grade-simplifier): 복잡한 텍스트를 어린 학습자가 이해하기 쉽게 만듭니다.
- [직유법의 달인](https://docs.anthropic.com/ko/prompt-library/simile-savant): 기본적인 설명으로부터 직유법을 생성합니다.
- [소크라테스식 현자](https://docs.anthropic.com/ko/prompt-library/socratic-sage): 사용자가 제시한 주제에 대해 소크라테스식 대화를 나눕니다.
- [스프레드시트 마법사](https://docs.anthropic.com/ko/prompt-library/spreadsheet-sorcerer): 다양한 유형의 데이터가 포함된 CSV 스프레드시트를 생성합니다.
- [SQL 마법사](https://docs.anthropic.com/ko/prompt-library/sql-sorcerer): 일상적인 언어를 SQL 쿼리로 변환합니다.
- [스토리텔링 파트너](https://docs.anthropic.com/ko/prompt-library/storytelling-sidekick): 사용자와 함께 흥미진진한 이야기를 만들고, 흥미로운 반전과 캐릭터 발전을 제안합니다.
- [시간 여행 컨설턴트](https://docs.anthropic.com/ko/prompt-library/time-travel-consultant): 사용자가 가상의 시간 여행 시나리오와 그 영향을 탐색하도록 돕습니다.
- [발음 연습 문장](https://docs.anthropic.com/ko/prompt-library/tongue-twister): 도전적인 발음 연습 문장을 만듭니다.
- [퀴즈 생성기](https://docs.anthropic.com/ko/prompt-library/trivia-generator): 다양한 주제에 대한 퀴즈 문제를 생성하고 필요할 때 힌트를 제공합니다.
- [트윗 톤 감지기](https://docs.anthropic.com/ko/prompt-library/tweet-tone-detector): 트윗 뒤에 숨겨진 톤과 감정을 감지합니다.
- [VR 피트니스 혁신가](https://docs.anthropic.com/ko/prompt-library/vr-fitness-innovator): 가상현실 피트니스 게임을 위한 창의적인 아이디어를 구상합니다.
- [웹사이트 마법사](https://docs.anthropic.com/ko/prompt-library/website-wizard): 사용자 요구사항에 따른 단일 페이지 웹사이트 제작
- [API](https://docs.anthropic.com/ko/release-notes/api): Anthropic의 API와 개발자 콘솔의 업데이트를 확인하세요.
- [Claude 앱](https://docs.anthropic.com/ko/release-notes/claude-apps): Anthropic의 Claude 애플리케이션 전반의 업데이트 소식을 확인하세요.
- [개요](https://docs.anthropic.com/ko/release-notes/overview): Anthropic의 제품과 서비스의 업데이트 소식을 확인하세요.
- [시스템 프롬프트](https://docs.anthropic.com/ko/release-notes/system-prompts): [Claude.ai](https://www.claude.ai)와 Claude [iOS](http://anthropic.com/ios) 및 [Android](http://anthropic.com/android) 앱의 핵심 시스템 프롬프트 업데이트를 확인하세요.
- [Cancelar um Lote de Mensagens (beta)](https://docs.anthropic.com/pt/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [API Amazon Bedrock](https://docs.anthropic.com/pt/api/claude-on-amazon-bedrock): Os modelos Claude da Anthropic agora estão geralmente disponíveis através do Amazon Bedrock.
- [API do Vertex AI](https://docs.anthropic.com/pt/api/claude-on-vertex-ai): Os modelos Claude da Anthropic agora estão geralmente disponíveis através do [Vertex AI](https://cloud.google.com/vertex-ai).
- [SDKs do Cliente](https://docs.anthropic.com/pt/api/client-sdks): Fornecemos bibliotecas em Python e TypeScript que facilitam o trabalho com a API da Anthropic.
- [Create a Text Completion](https://docs.anthropic.com/pt/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Criar um Lote de Mensagens (beta)](https://docs.anthropic.com/pt/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Erros](https://docs.anthropic.com/pt/api/errors)
- [Obtendo ajuda](https://docs.anthropic.com/pt/api/getting-help): Tentamos fornecer respostas para as perguntas mais comuns nesta documentação. No entanto, se você precisar de suporte técnico adicional usando Claude, a API da Anthropic ou qualquer um de nossos produtos, você pode entrar em contato com nossa equipe de suporte em [support.anthropic.com](https://support.anthropic.com).
- [Começando](https://docs.anthropic.com/pt/api/getting-started)
- [Endereços IP](https://docs.anthropic.com/pt/api/ip-addresses): Os serviços da Anthropic estão em um intervalo fixo de endereços IP. Você pode adicionar estes ao seu firewall para abrir a área mínima de superfície para tráfego de saída ao acessar a API e Console da Anthropic. Estes intervalos não serão alterados sem aviso prévio.
- [Listar Lotes de Mensagens (beta)](https://docs.anthropic.com/pt/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/pt/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Exemplos de Lotes de Mensagens](https://docs.anthropic.com/pt/api/messages-batch-examples): Exemplos de uso da API de Lotes de Mensagens
- [Contagem de tokens de Mensagem (beta)](https://docs.anthropic.com/pt/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Exemplos de Mensagens](https://docs.anthropic.com/pt/api/messages-examples): Exemplos de requisição e resposta para a API de Mensagens
- [Transmissão de Mensagens](https://docs.anthropic.com/pt/api/messages-streaming)
- [Migrando de Completions de Texto](https://docs.anthropic.com/pt/api/migrating-from-text-completions-to-messages): Migrando de Completions de Texto para Messages
- [Validação de prompt](https://docs.anthropic.com/pt/api/prompt-validation): Com Text Completions
- [Limites de taxa](https://docs.anthropic.com/pt/api/rate-limits): Para mitigar contra o uso indevido e gerenciar a capacidade em nossa API, implementamos limites sobre quanto uma organização pode usar a API do Claude.
- [Recuperar Resultados do Lote de Mensagens (beta)](https://docs.anthropic.com/pt/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Recuperar um Lote de Mensagens (beta)](https://docs.anthropic.com/pt/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Streaming de Conclusões de Texto](https://docs.anthropic.com/pt/api/streaming)
- [Regiões suportadas](https://docs.anthropic.com/pt/api/supported-regions): Aqui estão os países, regiões e territórios dos quais atualmente podemos suportar o acesso:
- [Versões](https://docs.anthropic.com/pt/api/versioning): Ao fazer solicitações à API, você deve enviar um cabeçalho de solicitação `anthropic-version`. Por exemplo, `anthropic-version: 2023-06-01`. Se você estiver usando nossas [bibliotecas cliente](/pt/api/client-libraries), isso é tratado automaticamente para você.
- [Atualizações de Julho de 2024](https://docs.anthropic.com/pt/developer-newsletter/july2024): 28 de julho de 2024
- [Visão Geral](https://docs.anthropic.com/pt/developer-newsletter/overview): Explore atualizações mensais, análises técnicas aprofundadas, melhores práticas e histórias de sucesso para aprimorar suas integrações com Claude
- [Modelos](https://docs.anthropic.com/pt/docs/about-claude/models): Claude é uma família de modelos de linguagem de última geração desenvolvidos pela Anthropic. Este guia apresenta nossos modelos e compara seu desempenho com modelos legados.
- [Segurança e conformidade](https://docs.anthropic.com/pt/docs/about-claude/security-compliance)
- [Moderação de conteúdo](https://docs.anthropic.com/pt/docs/about-claude/use-case-guides/content-moderation): A moderação de conteúdo é um aspecto crítico para manter um ambiente seguro, respeitoso e produtivo em aplicações digitais. Neste guia, discutiremos como o Claude pode ser usado para moderar conteúdo dentro da sua aplicação digital.
- [Agente de suporte ao cliente](https://docs.anthropic.com/pt/docs/about-claude/use-case-guides/customer-support-chat): Este guia mostra como aproveitar as capacidades avançadas de conversação do Claude para lidar com consultas de clientes em tempo real, fornecendo suporte 24/7, reduzindo tempos de espera e gerenciando altos volumes de suporte com respostas precisas e interações positivas.
- [Sumarização legal](https://docs.anthropic.com/pt/docs/about-claude/use-case-guides/legal-summarization): Este guia mostra como aproveitar as capacidades avançadas de processamento de linguagem natural do Claude para sumarizar documentos legais de forma eficiente, extraindo informações-chave e acelerando a pesquisa jurídica. Com o Claude, você pode otimizar a revisão de contratos, preparação de litígios e trabalho regulatório, economizando tempo e garantindo precisão em seus processos legais.
- [Guias para casos de uso comuns](https://docs.anthropic.com/pt/docs/about-claude/use-case-guides/overview)
- [Roteamento de tickets](https://docs.anthropic.com/pt/docs/about-claude/use-case-guides/ticket-routing): Este guia mostra como aproveitar as capacidades avançadas de compreensão de linguagem natural do Claude para classificar tickets de suporte ao cliente em escala com base na intenção do cliente, urgência, priorização, perfil do cliente e muito mais.
- [Extensão do Google Sheets](https://docs.anthropic.com/pt/docs/build-with-claude/claude-for-sheets): A [extensão Claude for Sheets](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) integra o Claude ao Google Sheets, permitindo executar interações com o Claude diretamente nas células.
- [null](https://docs.anthropic.com/pt/docs/build-with-claude/computer-use)
- [Defina seus critérios de sucesso](https://docs.anthropic.com/pt/docs/build-with-claude/define-success)
- [Criar avaliações empíricas fortes](https://docs.anthropic.com/pt/docs/build-with-claude/develop-tests)
- [Embeddings](https://docs.anthropic.com/pt/docs/build-with-claude/embeddings): Embeddings de texto são representações numéricas de texto que permitem medir a similaridade semântica. Este guia introduz embeddings, suas aplicações e como usar modelos de embedding para tarefas como busca, recomendações e detecção de anomalias.
- [Lotes de Mensagens (beta)](https://docs.anthropic.com/pt/docs/build-with-claude/message-batches)
- [Suporte a PDF (beta)](https://docs.anthropic.com/pt/docs/build-with-claude/pdf-support): Os modelos Claude 3.5 Sonnet agora suportam entrada de PDF e compreendem tanto o conteúdo textual quanto visual dentro dos documentos.
- [Cache de prompts (beta)](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-caching)
- [Seja claro, direto e detalhado](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Deixe o Claude pensar (prompting de cadeia de pensamento) para aumentar o desempenho](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Encadeie prompts complexos para um desempenho mais forte](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Dicas para prompts com contexto longo](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Use exemplos (prompting multishot) para guiar o comportamento do Claude](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Visão geral da engenharia de prompts](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/overview)
- [Pré-preencher a resposta do Claude para maior controle de saída](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Gerar automaticamente modelos de prompt iniciais](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Use nosso aprimorador de prompts para otimizar seus prompts](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Use modelos de prompt e variáveis](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Dando a Claude um papel com um prompt de sistema](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/system-prompts)
- [Use tags XML para estruturar seus prompts](https://docs.anthropic.com/pt/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Geração de texto](https://docs.anthropic.com/pt/docs/build-with-claude/text-generation): O Claude se destaca em uma ampla gama de tarefas baseadas em texto. O Claude foi treinado para processar código, prosa e outras entradas em linguagem natural, e fornecer saídas de texto em resposta.
- [Contagem de tokens (beta)](https://docs.anthropic.com/pt/docs/build-with-claude/token-counting)
- [Uso de ferramentas (chamada de função)](https://docs.anthropic.com/pt/docs/build-with-claude/tool-use)
- [Visão](https://docs.anthropic.com/pt/docs/build-with-claude/vision): A família de modelos Claude 3 vem com novas capacidades de visão que permitem ao Claude entender e analisar imagens, abrindo possibilidades empolgantes para interação multimodal.
- [Configuração inicial](https://docs.anthropic.com/pt/docs/initial-setup): Vamos aprender como usar a API da Anthropic para construir com Claude.
- [Introdução ao Claude](https://docs.anthropic.com/pt/docs/intro-to-claude): Claude é uma família de [modelos de IA altamente performáticos e inteligentes](/pt/docs/about-claude/models) construídos pela Anthropic. Embora o Claude seja poderoso e extensível, ele também é a IA mais confiável e segura disponível. Ele segue protocolos críticos, comete menos erros e é resistente a violações—permitindo que [clientes empresariais](https://www.anthropic.com/customers) construam as aplicações mais seguras alimentadas por IA em escala.
- [Política de Privacidade da Anthropic](https://docs.anthropic.com/pt/docs/legal-center/privacy)
- [Livro de Receitas Anthropic](https://docs.anthropic.com/pt/docs/resources/cookbook)
- [Cursos Anthropic](https://docs.anthropic.com/pt/docs/resources/courses)
- [Glossário](https://docs.anthropic.com/pt/docs/resources/glossary): Estes conceitos não são exclusivos dos modelos de linguagem da Anthropic, mas apresentamos um breve resumo dos termos-chave abaixo.
- [Cartão do modelo Claude 3](https://docs.anthropic.com/pt/docs/resources/model-card)
- [Descontinuação de Modelos](https://docs.anthropic.com/pt/docs/resources/model-deprecations)
- [Status do sistema](https://docs.anthropic.com/pt/docs/resources/status)
- [Usando a Ferramenta de Avaliação](https://docs.anthropic.com/pt/docs/test-and-evaluate/eval-tool): O [Anthropic Console](https://console.anthropic.com/dashboard) possui uma **ferramenta de Avaliação** que permite testar seus prompts em vários cenários.
- [Aumentar a consistência da saída (modo JSON)](https://docs.anthropic.com/pt/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Mantenha o Claude no personagem com prompts de papel e pré-preenchimento](https://docs.anthropic.com/pt/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Mitigar jailbreaks e injeções de prompt](https://docs.anthropic.com/pt/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Reduzir alucinações](https://docs.anthropic.com/pt/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Reduzindo a latência](https://docs.anthropic.com/pt/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Reduzir vazamento de prompt](https://docs.anthropic.com/pt/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Bem-vindo ao Claude](https://docs.anthropic.com/pt/docs/welcome): Claude é uma plataforma de IA altamente performática, confiável e inteligente construída pela Anthropic. Claude se destaca em tarefas envolvendo linguagem, raciocínio, análise, programação e muito mais.
- [null](https://docs.anthropic.com/pt/home)
- [Editor adaptativo](https://docs.anthropic.com/pt/prompt-library/adaptive-editor): Reescreva texto seguindo instruções do usuário, como com um tom, público ou estilo diferente.
- [Analisador de códigos de aeroporto](https://docs.anthropic.com/pt/prompt-library/airport-code-analyst): Encontrar e extrair códigos de aeroporto do texto.
- [Antropólogo alienígena](https://docs.anthropic.com/pt/prompt-library/alien-anthropologist): Analise a cultura e os costumes humanos da perspectiva de um antropólogo alienígena.
- [Alquimista da aliteração](https://docs.anthropic.com/pt/prompt-library/alliteration-alchemist): Gere frases e sentenças aliterativas para qualquer assunto.
- [Transmissões de Babel](https://docs.anthropic.com/pt/prompt-library/babels-broadcasts): Crie tweets convincentes de anúncios de produtos nos 10 idiomas mais falados do mundo.
- [Construtor de marca](https://docs.anthropic.com/pt/prompt-library/brand-builder): Crie um briefing de design para uma identidade de marca holística.
- [Coach de carreira](https://docs.anthropic.com/pt/prompt-library/career-coach): Participe de conversas de role-play com um coach de carreira com IA.
- [Cite suas fontes](https://docs.anthropic.com/pt/prompt-library/cite-your-sources): Obtenha respostas para perguntas sobre o conteúdo de um documento com citações relevantes que apoiam a resposta.
- [Esclarecedor de código](https://docs.anthropic.com/pt/prompt-library/code-clarifier): Simplifica e explica código complexo em linguagem simples.
- [Consultor de código](https://docs.anthropic.com/pt/prompt-library/code-consultant): Sugere melhorias para otimizar o desempenho do código Python.
- [Clarividente corporativo](https://docs.anthropic.com/pt/prompt-library/corporate-clairvoyant): Extraia insights, identifique riscos e destile informações-chave de longos relatórios corporativos em um único memorando.
- [Teclas Cósmicas](https://docs.anthropic.com/pt/prompt-library/cosmic-keystrokes): Gere um jogo interativo de digitação rápida em um único arquivo HTML, com jogabilidade de rolagem lateral e estilização Tailwind CSS.
- [Conversor CSV](https://docs.anthropic.com/pt/prompt-library/csv-converter): Converta dados de vários formatos (JSON, XML, etc.) em arquivos CSV corretamente formatados.
- [Criador culinário](https://docs.anthropic.com/pt/prompt-library/culinary-creator): Sugere ideias de receitas com base nos ingredientes disponíveis e preferências alimentares do usuário.
- [Organizador de dados](https://docs.anthropic.com/pt/prompt-library/data-organizer): Transforme texto não estruturado em tabelas JSON personalizadas.
- [Decodificador de direções](https://docs.anthropic.com/pt/prompt-library/direction-decoder): Transforme linguagem natural em instruções passo a passo.
- [Intérprete de sonhos](https://docs.anthropic.com/pt/prompt-library/dream-interpreter): Oferece interpretações e insights sobre o simbolismo dos sonhos do usuário
- [Estimador de eficiência](https://docs.anthropic.com/pt/prompt-library/efficiency-estimator): Calcule a complexidade temporal de funções e algoritmos.
- [Extrator de email](https://docs.anthropic.com/pt/prompt-library/email-extractor): Extrair endereços de email de um documento em uma lista formatada em JSON.
- [Codificador de emoji](https://docs.anthropic.com/pt/prompt-library/emoji-encoder): Converta texto simples em mensagens divertidas e expressivas com emojis.
- [Navegador de dilemas éticos](https://docs.anthropic.com/pt/prompt-library/ethical-dilemma-navigator): Ajude o usuário a pensar em dilemas éticos complexos e forneça diferentes perspectivas.
- [Especialista em fórmulas do Excel](https://docs.anthropic.com/pt/prompt-library/excel-formula-expert): Crie fórmulas do Excel baseadas em cálculos ou manipulações de dados descritos pelo usuário.
- [Fabricador de funções](https://docs.anthropic.com/pt/prompt-library/function-fabricator): Crie funções Python baseadas em especificações detalhadas.
- [Consultor de moda futurista](https://docs.anthropic.com/pt/prompt-library/futuristic-fashion-advisor): Sugere tendências e estilos de moda vanguardistas baseados nas preferências específicas do usuário.
- [Domine o Git](https://docs.anthropic.com/pt/prompt-library/git-gud): Gere comandos Git apropriados com base em ações de controle de versão descritas pelo usuário.
- [Programador de scripts do Google Apps](https://docs.anthropic.com/pt/prompt-library/google-apps-scripter): Gere scripts do Google Apps para completar tarefas com base nos requisitos do usuário.
- [Guru da avaliação](https://docs.anthropic.com/pt/prompt-library/grading-guru): Compare e avalie a qualidade de textos escritos com base em critérios e padrões definidos pelo usuário.
- [Gênio da gramática](https://docs.anthropic.com/pt/prompt-library/grammar-genie): Transforme frases gramaticalmente incorretas em inglês adequado.
- [Hal, o ajudante bem-humorado](https://docs.anthropic.com/pt/prompt-library/hal-the-humorous-helper): Converse com uma IA experiente que tem um lado sarcástico.
- [Iluminador de expressões idiomáticas](https://docs.anthropic.com/pt/prompt-library/idiom-illuminator): Explique o significado e a origem de expressões idiomáticas e provérbios comuns.
- [Gerador de perguntas para entrevistas](https://docs.anthropic.com/pt/prompt-library/interview-question-crafter): Gere perguntas para entrevistas.
- [Legenda LaTeX](https://docs.anthropic.com/pt/prompt-library/latex-legend): Escreva documentos LaTeX, gerando código para equações matemáticas, tabelas e mais.
- [Planejador de aulas](https://docs.anthropic.com/pt/prompt-library/lesson-planner): Crie planos de aula detalhados sobre qualquer assunto.
- [Biblioteca](https://docs.anthropic.com/pt/prompt-library/library)
- [Moderador mestre](https://docs.anthropic.com/pt/prompt-library/master-moderator): Avalie entradas do usuário quanto a conteúdo potencialmente prejudicial ou ilegal.
- [Escriba de reunião](https://docs.anthropic.com/pt/prompt-library/meeting-scribe): Destile reuniões em resumos concisos incluindo tópicos de discussão, principais conclusões e itens de ação.
- [Mestre dos memorandos](https://docs.anthropic.com/pt/prompt-library/memo-maestro): Componha memorandos empresariais abrangentes com base em pontos-chave.
- [Mentor de mindfulness](https://docs.anthropic.com/pt/prompt-library/mindfulness-mentor): Guie o usuário através de exercícios e técnicas de mindfulness para redução do estresse.
- [Coloridor de humor](https://docs.anthropic.com/pt/prompt-library/mood-colorizer): Transforme descrições textuais de humores em códigos HEX correspondentes.
- [Musa motivacional](https://docs.anthropic.com/pt/prompt-library/motivational-muse): Fornecer mensagens motivacionais e afirmações personalizadas com base na entrada do usuário
- [Criador de Neologismos](https://docs.anthropic.com/pt/prompt-library/neologism-creator): Invente novas palavras e forneça suas definições com base em conceitos ou ideias fornecidos pelo usuário.
- [Ponderador de perspectivas](https://docs.anthropic.com/pt/prompt-library/perspectives-ponderer): Avalie os prós e contras de um tópico fornecido pelo usuário.
- [Reflexões filosóficas](https://docs.anthropic.com/pt/prompt-library/philosophical-musings): Participe de discussões filosóficas profundas e experimentos mentais.
- [Purificador de PII](https://docs.anthropic.com/pt/prompt-library/pii-purifier): Detecte e remova automaticamente informações pessoalmente identificáveis (PII) de documentos de texto.
- [Superpoderes poliglotas](https://docs.anthropic.com/pt/prompt-library/polyglot-superpowers): Traduza texto de qualquer idioma para qualquer idioma.
- [Poeta de palavras-valise](https://docs.anthropic.com/pt/prompt-library/portmanteau-poet): Combine duas palavras para criar uma nova palavra-valise significativa.
- [Profissional de nomes de produtos](https://docs.anthropic.com/pt/prompt-library/product-naming-pro): Crie nomes chamativos para produtos a partir de descrições e palavras-chave
- [Polidor de prosa](https://docs.anthropic.com/pt/prompt-library/prose-polisher): Refine e melhore conteúdo escrito com técnicas e sugestões avançadas de revisão.
- [Trocadilhista](https://docs.anthropic.com/pt/prompt-library/pun-dit): Gere trocadilhos e jogos de palavras inteligentes baseados em qualquer tópico.
- [Caçador de bugs em Python](https://docs.anthropic.com/pt/prompt-library/python-bug-buster): Detecte e corrija bugs em código Python.
- [Classificador de avaliações](https://docs.anthropic.com/pt/prompt-library/review-classifier): Categorize feedback em tags e categorizações pré-especificadas.
- [Me decifre](https://docs.anthropic.com/pt/prompt-library/riddle-me-this): Gere enigmas e guie o usuário até as soluções.
- [Simulador de cenários de ficção científica](https://docs.anthropic.com/pt/prompt-library/sci-fi-scenario-simulator): Discuta com o usuário vários cenários de ficção científica e os desafios e considerações associados.
- [Simplificador para segunda série](https://docs.anthropic.com/pt/prompt-library/second-grade-simplifier): Torne textos complexos fáceis de entender para jovens aprendizes
- [Mestre das Símiles](https://docs.anthropic.com/pt/prompt-library/simile-savant): Gere símiles a partir de descrições básicas.
- [Sábio socrático](https://docs.anthropic.com/pt/prompt-library/socratic-sage): Participe de uma conversa no estilo socrático sobre um tópico fornecido pelo usuário
- [Feiticeiro de planilhas](https://docs.anthropic.com/pt/prompt-library/spreadsheet-sorcerer): Gere planilhas CSV com vários tipos de dados.
- [Feiticeiro SQL](https://docs.anthropic.com/pt/prompt-library/sql-sorcerer): Transforme linguagem cotidiana em consultas SQL.
- [Companheiro de narrativas](https://docs.anthropic.com/pt/prompt-library/storytelling-sidekick): Crie histórias envolventes em colaboração com o usuário, oferecendo reviravoltas no enredo e desenvolvimento de personagens.
- [Consultor de viagem no tempo](https://docs.anthropic.com/pt/prompt-library/time-travel-consultant): Ajude o usuário a navegar por cenários hipotéticos de viagem no tempo e suas implicações.
- [Trava-línguas](https://docs.anthropic.com/pt/prompt-library/tongue-twister): Crie trava-línguas desafiadores.
- [Gerador de trivia](https://docs.anthropic.com/pt/prompt-library/trivia-generator): Gere perguntas de trivia sobre uma ampla variedade de tópicos e forneça dicas quando necessário.
- [Detector de tom de tweets](https://docs.anthropic.com/pt/prompt-library/tweet-tone-detector): Detecte o tom e o sentimento por trás dos tweets.
- [Inovador de fitness em RV](https://docs.anthropic.com/pt/prompt-library/vr-fitness-innovator): Gerar ideias criativas para jogos de fitness em realidade virtual.
- [Assistente de website](https://docs.anthropic.com/pt/prompt-library/website-wizard): Crie websites de uma página baseados em especificações do usuário.
- [API](https://docs.anthropic.com/pt/release-notes/api): Acompanhe as atualizações da API da Anthropic e do Console do Desenvolvedor.
- [Aplicativos Claude](https://docs.anthropic.com/pt/release-notes/claude-apps): Acompanhe as atualizações dos aplicativos Claude da Anthropic.
- [Visão Geral](https://docs.anthropic.com/pt/release-notes/overview): Acompanhe as atualizações dos produtos e serviços da Anthropic.
- [Prompts do Sistema](https://docs.anthropic.com/pt/release-notes/system-prompts): Veja atualizações dos prompts principais do sistema no [Claude.ai](https://www.claude.ai) e nos aplicativos do Claude para [iOS](http://anthropic.com/ios) e [Android](http://anthropic.com/android).
- [Отмена пакета сообщений (бета)](https://docs.anthropic.com/ru/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/ru/api/claude-on-amazon-bedrock): Модели Claude от Anthropic теперь общедоступны через Amazon Bedrock.
- [Vertex AI API](https://docs.anthropic.com/ru/api/claude-on-vertex-ai): Модели Claude от Anthropic теперь общедоступны через [Vertex AI](https://cloud.google.com/vertex-ai).
- [Клиентские SDK](https://docs.anthropic.com/ru/api/client-sdks): Мы предоставляем библиотеки на Python и TypeScript, которые упрощают работу с API Anthropic.
- [Create a Text Completion](https://docs.anthropic.com/ru/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [Создание пакета сообщений (бета)](https://docs.anthropic.com/ru/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [Ошибки](https://docs.anthropic.com/ru/api/errors)
- [Получение помощи](https://docs.anthropic.com/ru/api/getting-help): Мы постарались предоставить ответы на наиболее распространенные вопросы в этой документации. Однако, если вам нужна дополнительная техническая поддержка по использованию Claude, API Anthropic или любого из наших продуктов, вы можете обратиться к нашей команде поддержки на [support.anthropic.com](https://support.anthropic.com).
- [Начало работы](https://docs.anthropic.com/ru/api/getting-started)
- [IP-адреса](https://docs.anthropic.com/ru/api/ip-addresses): Сервисы Anthropic работают в фиксированном диапазоне IP-адресов. Вы можете добавить их в свой файрвол, чтобы открыть минимальную поверхность для исходящего трафика при доступе к API и Консоли Anthropic. Эти диапазоны не будут изменяться без уведомления.
- [Список пакетов сообщений (бета)](https://docs.anthropic.com/ru/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/ru/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [Примеры пакетной обработки сообщений](https://docs.anthropic.com/ru/api/messages-batch-examples): Примеры использования API пакетной обработки сообщений
- [Подсчет токенов сообщения (бета)](https://docs.anthropic.com/ru/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [Примеры сообщений](https://docs.anthropic.com/ru/api/messages-examples): Примеры запросов и ответов для API сообщений
- [Потоковая передача сообщений](https://docs.anthropic.com/ru/api/messages-streaming)
- [Миграция с Text Completions](https://docs.anthropic.com/ru/api/migrating-from-text-completions-to-messages): Миграция с Text Completions на Messages
- [Валидация промпта](https://docs.anthropic.com/ru/api/prompt-validation): С помощью Text Completions
- [Лимиты использования](https://docs.anthropic.com/ru/api/rate-limits): Чтобы предотвратить злоупотребления и управлять нагрузкой на наш API, мы установили ограничения на то, как много организация может использовать API Claude.
- [Получение результатов пакетной обработки сообщений (бета)](https://docs.anthropic.com/ru/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [Получить пакет сообщений (бета)](https://docs.anthropic.com/ru/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [Потоковая передача текстовых дополнений](https://docs.anthropic.com/ru/api/streaming)
- [Поддерживаемые регионы](https://docs.anthropic.com/ru/api/supported-regions): Вот страны, регионы и территории, из которых мы в настоящее время поддерживаем доступ:
- [Версии](https://docs.anthropic.com/ru/api/versioning): При выполнении API-запросов необходимо отправлять заголовок запроса `anthropic-version`. Например, `anthropic-version: 2023-06-01`. Если вы используете наши [клиентские библиотеки](/ru/api/client-libraries), это обрабатывается автоматически.
- [Обновления за июль 2024](https://docs.anthropic.com/ru/developer-newsletter/july2024): 28 июля 2024
- [Обзор](https://docs.anthropic.com/ru/developer-newsletter/overview): Изучите ежемесячные обновления, технические обзоры, лучшие практики и истории успеха для улучшения ваших интеграций с Claude
- [Модели](https://docs.anthropic.com/ru/docs/about-claude/models): Claude - это семейство передовых больших языковых моделей, разработанных компанией Anthropic. Это руководство представляет наши модели и сравнивает их производительность с предыдущими моделями.
- [Безопасность и соответствие требованиям](https://docs.anthropic.com/ru/docs/about-claude/security-compliance)
- [Модерация контента](https://docs.anthropic.com/ru/docs/about-claude/use-case-guides/content-moderation): Модерация контента является критически важным аспектом поддержания безопасной, уважительной и продуктивной среды в цифровых приложениях. В этом руководстве мы обсудим, как Claude можно использовать для модерации контента в вашем цифровом приложении.
- [Агент службы поддержки клиентов](https://docs.anthropic.com/ru/docs/about-claude/use-case-guides/customer-support-chat): Это руководство показывает, как использовать продвинутые разговорные возможности Claude для обработки запросов клиентов в реальном времени, обеспечивая круглосуточную поддержку, сокращая время ожидания и управляя большим объемом обращений с точными ответами и позитивным взаимодействием.
- [Юридическое резюмирование](https://docs.anthropic.com/ru/docs/about-claude/use-case-guides/legal-summarization): Это руководство показывает, как использовать продвинутые возможности обработки естественного языка Claude для эффективного резюмирования юридических документов, извлечения ключевой информации и ускорения юридических исследований. С помощью Claude вы можете оптимизировать проверку контрактов, подготовку к судебным разбирательствам и нормативную работу, экономя время и обеспечивая точность в ваших юридических процессах.
- [Руководства по распространенным сценариям использования](https://docs.anthropic.com/ru/docs/about-claude/use-case-guides/overview)
- [Маршрутизация тикетов](https://docs.anthropic.com/ru/docs/about-claude/use-case-guides/ticket-routing): Это руководство объясняет, как использовать продвинутые возможности понимания естественного языка Claude для классификации тикетов службы поддержки в масштабе на основе намерений клиентов, срочности, приоритизации, профиля клиента и других параметров.
- [Дополнение для Google Sheets](https://docs.anthropic.com/ru/docs/build-with-claude/claude-for-sheets): [Расширение Claude для Sheets](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257) интегрирует Claude в Google Sheets, позволяя выполнять взаимодействие с Claude непосредственно в ячейках.
- [Использование компьютера (бета)](https://docs.anthropic.com/ru/docs/build-with-claude/computer-use): Обновленная модель Claude 3.5 Sonnet способна взаимодействовать с [инструментами](/ru/docs/build-with-claude/tool-use), которые могут манипулировать средой рабочего стола компьютера.
- [Определите критерии успеха](https://docs.anthropic.com/ru/docs/build-with-claude/define-success)
- [Создание эффективных эмпирических оценок](https://docs.anthropic.com/ru/docs/build-with-claude/develop-tests)
- [Эмбеддинги](https://docs.anthropic.com/ru/docs/build-with-claude/embeddings): Текстовые эмбеддинги - это числовые представления текста, позволяющие измерять семантическое сходство. Это руководство знакомит с эмбеддингами, их применением и использованием моделей эмбеддингов для таких задач, как поиск, рекомендации и обнаружение аномалий.
- [Пакетная обработка сообщений (бета)](https://docs.anthropic.com/ru/docs/build-with-claude/message-batches)
- [Поддержка PDF (бета)](https://docs.anthropic.com/ru/docs/build-with-claude/pdf-support): Модели Claude 3.5 Sonnet теперь поддерживают ввод PDF и понимают как текстовое, так и визуальное содержимое документов.
- [Кэширование промптов (бета)](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-caching)
- [Будьте ясными, прямыми и подробными](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [Позвольте Claude думать (цепочка рассуждений) для повышения производительности](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [Цепочка сложных промптов для повышения производительности](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/chain-prompts)
- [Советы по работе с длинным контекстом](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/long-context-tips)
- [Используйте примеры (мультишотовый промптинг) для управления поведением Claude](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [Обзор инженерии промптов](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/overview)
- [Предварительное заполнение ответа Claude для лучшего контроля вывода](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [Автоматическая генерация черновых шаблонов промптов](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/prompt-generator)
- [Используйте наш улучшатель промптов для оптимизации ваших промптов](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/prompt-improver)
- [Использование шаблонов промптов и переменных](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [Назначение роли Claude с помощью системного промпта](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/system-prompts)
- [Используйте XML-теги для структурирования ваших промптов](https://docs.anthropic.com/ru/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [Генерация текста](https://docs.anthropic.com/ru/docs/build-with-claude/text-generation): Claude превосходно справляется с широким спектром текстовых задач. Claude обучен обрабатывать код, прозу и другие естественные языковые данные и предоставлять текстовые ответы.
- [Подсчет токенов (бета)](https://docs.anthropic.com/ru/docs/build-with-claude/token-counting)
- [Использование инструментов (вызов функций)](https://docs.anthropic.com/ru/docs/build-with-claude/tool-use)
- [Зрение](https://docs.anthropic.com/ru/docs/build-with-claude/vision): Семейство моделей Claude 3 обладает новыми возможностями зрительного восприятия, которые позволяют Claude понимать и анализировать изображения, открывая захватывающие возможности для мультимодального взаимодействия.
- [Начальная настройка](https://docs.anthropic.com/ru/docs/initial-setup): Давайте научимся использовать API Anthropic для работы с Claude.
- [Введение в Claude](https://docs.anthropic.com/ru/docs/intro-to-claude): Claude - это семейство [высокопроизводительных и интеллектуальных ИИ-моделей](/ru/docs/about-claude/models), созданных компанией Anthropic. Хотя Claude является мощным и расширяемым, это также самый надежный и достоверный ИИ из доступных. Он следует критически важным протоколам, делает меньше ошибок и устойчив к взломам, позволяя [корпоративным клиентам](https://www.anthropic.com/customers) создавать самые безопасные ИИ-приложения в масштабе.
- [Политика конфиденциальности Anthropic](https://docs.anthropic.com/ru/docs/legal-center/privacy)
- [Сборник рецептов Anthropic](https://docs.anthropic.com/ru/docs/resources/cookbook)
- [Курсы Anthropic](https://docs.anthropic.com/ru/docs/resources/courses)
- [Глоссарий](https://docs.anthropic.com/ru/docs/resources/glossary): Эти концепции не уникальны для языковых моделей Anthropic, но мы представляем краткое изложение ключевых терминов ниже.
- [Карточка модели Claude 3](https://docs.anthropic.com/ru/docs/resources/model-card)
- [Прекращение поддержки моделей](https://docs.anthropic.com/ru/docs/resources/model-deprecations)
- [Статус системы](https://docs.anthropic.com/ru/docs/resources/status)
- [Использование инструмента оценки](https://docs.anthropic.com/ru/docs/test-and-evaluate/eval-tool): В [Anthropic Console](https://console.anthropic.com/dashboard) есть **инструмент оценки**, который позволяет тестировать ваши промпты в различных сценариях.
- [Повышение согласованности вывода (режим JSON)](https://docs.anthropic.com/ru/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [Поддержание характера Claude с помощью ролевых промптов и предварительного заполнения](https://docs.anthropic.com/ru/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [Предотвращение взлома и инъекций промптов](https://docs.anthropic.com/ru/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [Уменьшение галлюцинаций](https://docs.anthropic.com/ru/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [Снижение задержки](https://docs.anthropic.com/ru/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [Уменьшение утечки промпта](https://docs.anthropic.com/ru/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [Добро пожаловать в Claude](https://docs.anthropic.com/ru/docs/welcome): Claude - это высокопроизводительная, надежная и интеллектуальная платформа искусственного интеллекта, созданная Anthropic. Claude превосходно справляется с задачами, связанными с языком, рассуждениями, анализом, программированием и многим другим.
- [null](https://docs.anthropic.com/ru/home)
- [Адаптивный редактор](https://docs.anthropic.com/ru/prompt-library/adaptive-editor): Переписывание текста в соответствии с инструкциями пользователя, например, с другим тоном, аудиторией или стилем.
- [Анализатор кодов аэропортов](https://docs.anthropic.com/ru/prompt-library/airport-code-analyst): Поиск и извлечение кодов аэропортов из текста
- [Антрополог-инопланетянин](https://docs.anthropic.com/ru/prompt-library/alien-anthropologist): Анализ человеческой культуры и обычаев с точки зрения инопланетного антрополога
- [Мастер аллитерации](https://docs.anthropic.com/ru/prompt-library/alliteration-alchemist): Создание аллитерационных фраз и предложений для любой заданной темы.
- [Трансляции Вавилона](https://docs.anthropic.com/ru/prompt-library/babels-broadcasts): Создавайте убедительные твиты о продуктах на 10 самых распространенных языках мира
- [Создатель бренда](https://docs.anthropic.com/ru/prompt-library/brand-builder): Создание дизайн-брифа для целостной айдентики бренда
- [Карьерный коуч](https://docs.anthropic.com/ru/prompt-library/career-coach): Участвуйте в ролевых беседах с ИИ-карьерным коучем.
- [Цитируйте свои источники](https://docs.anthropic.com/ru/prompt-library/cite-your-sources): Получайте ответы на вопросы о содержании документа с соответствующими цитатами, подтверждающими ответ.
- [Пояснитель кода](https://docs.anthropic.com/ru/prompt-library/code-clarifier): Упрощение и объяснение сложного кода простым языком.
- [Консультант по коду](https://docs.anthropic.com/ru/prompt-library/code-consultant): Предлагает улучшения для оптимизации производительности Python-кода.
- [Корпоративный провидец](https://docs.anthropic.com/ru/prompt-library/corporate-clairvoyant): Извлечение инсайтов, выявление рисков и выделение ключевой информации из длинных корпоративных отчетов в единый меморандум.
- [Космические Клавиши](https://docs.anthropic.com/ru/prompt-library/cosmic-keystrokes): Создайте интерактивную игру для тренировки скорости печати в одном HTML-файле с боковой прокруткой и стилизацией Tailwind CSS.
- [Конвертер CSV](https://docs.anthropic.com/ru/prompt-library/csv-converter): Преобразование данных из различных форматов (JSON, XML и др.) в правильно отформатированные CSV-файлы.
- [Кулинарный создатель](https://docs.anthropic.com/ru/prompt-library/culinary-creator): Предлагает идеи рецептов на основе доступных ингредиентов и диетических предпочтений пользователя.
- [Организатор данных](https://docs.anthropic.com/ru/prompt-library/data-organizer): Преобразование неструктурированного текста в специализированные JSON-таблицы.
- [Декодер направлений](https://docs.anthropic.com/ru/prompt-library/direction-decoder): Преобразование естественного языка в пошаговые инструкции.
- [Толкователь снов](https://docs.anthropic.com/ru/prompt-library/dream-interpreter): Предлагает интерпретации и понимание символизма снов пользователя
- [Оценщик эффективности](https://docs.anthropic.com/ru/prompt-library/efficiency-estimator): Расчет временной сложности функций и алгоритмов
- [Извлечение электронных адресов](https://docs.anthropic.com/ru/prompt-library/email-extractor): Извлечение адресов электронной почты из документа в список в формате JSON.
- [Эмодзи кодировщик](https://docs.anthropic.com/ru/prompt-library/emoji-encoder): Преобразуйте обычный текст в веселые и выразительные сообщения с эмодзи
- [Навигатор по этическим дилеммам](https://docs.anthropic.com/ru/prompt-library/ethical-dilemma-navigator): Помощь пользователю в осмыслении сложных этических дилемм и предоставление различных точек зрения
- [Эксперт по формулам Excel](https://docs.anthropic.com/ru/prompt-library/excel-formula-expert): Создание формул Excel на основе вычислений или манипуляций с данными, описанных пользователем.
- [Создатель функций](https://docs.anthropic.com/ru/prompt-library/function-fabricator): Создание функций Python на основе подробных спецификаций
- [Футуристический консультант по моде](https://docs.anthropic.com/ru/prompt-library/futuristic-fashion-advisor): Предлагает авангардные модные тренды и стили в соответствии с конкретными предпочтениями пользователя
- [Git gud](https://docs.anthropic.com/ru/prompt-library/git-gud): Генерация соответствующих команд Git на основе описанных пользователем действий по контролю версий.
- [Разработчик скриптов Google Apps](https://docs.anthropic.com/ru/prompt-library/google-apps-scripter): Создание скриптов Google Apps для выполнения задач на основе требований пользователя
- [Эксперт по оценке](https://docs.anthropic.com/ru/prompt-library/grading-guru): Сравнение и оценка качества письменных текстов на основе определенных пользователем критериев и стандартов
- [Грамматический гений](https://docs.anthropic.com/ru/prompt-library/grammar-genie): Преобразование грамматически неправильных предложений в правильный английский язык.
- [Хэл - помощник с чувством юмора](https://docs.anthropic.com/ru/prompt-library/hal-the-humorous-helper): Общайтесь с эрудированным ИИ, у которого есть саркастическая сторона
- [Толкователь идиом](https://docs.anthropic.com/ru/prompt-library/idiom-illuminator): Объяснение значения и происхождения распространенных идиом и пословиц
- [Составитель вопросов для собеседования](https://docs.anthropic.com/ru/prompt-library/interview-question-crafter): Генерация вопросов для собеседований.
- [Справочник по LaTeX](https://docs.anthropic.com/ru/prompt-library/latex-legend): Создавайте документы LaTeX, генерируя код для математических уравнений, таблиц и многого другого.
- [Планировщик уроков](https://docs.anthropic.com/ru/prompt-library/lesson-planner): Создавайте подробные планы уроков по любому предмету.
- [Библиотека](https://docs.anthropic.com/ru/prompt-library/library)
- [Главный модератор](https://docs.anthropic.com/ru/prompt-library/master-moderator): Оценка пользовательских запросов на предмет потенциально вредного или незаконного содержания
- [Секретарь собрания](https://docs.anthropic.com/ru/prompt-library/meeting-scribe): Преобразование собраний в краткие сводки, включающие темы обсуждения, ключевые выводы и задачи.
- [Мастер служебных записок](https://docs.anthropic.com/ru/prompt-library/memo-maestro): Составление подробных служебных записок компании на основе ключевых пунктов
- [Наставник по осознанности](https://docs.anthropic.com/ru/prompt-library/mindfulness-mentor): Проведите пользователя через упражнения по осознанности и техники снижения стресса
- [Цветовой определитель настроения](https://docs.anthropic.com/ru/prompt-library/mood-colorizer): Преобразование текстовых описаний настроений в соответствующие HEX-коды.
- [Мотивационная муза](https://docs.anthropic.com/ru/prompt-library/motivational-muse): Предоставление персонализированных мотивационных сообщений и аффирмаций на основе пользовательского ввода
- [Создатель неологизмов](https://docs.anthropic.com/ru/prompt-library/neologism-creator): Изобретение новых слов и предоставление их определений на основе предоставленных пользователем концепций или идей.
- [Анализатор перспектив](https://docs.anthropic.com/ru/prompt-library/perspectives-ponderer): Взвешивание плюсов и минусов предложенной пользователем темы
- [Философские размышления](https://docs.anthropic.com/ru/prompt-library/philosophical-musings): Погружение в глубокие философские дискуссии и мысленные эксперименты
- [Очиститель персональных данных](https://docs.anthropic.com/ru/prompt-library/pii-purifier): Автоматическое обнаружение и удаление персональной идентифицирующей информации (PII) из текстовых документов.
- [Полиглотские суперспособности](https://docs.anthropic.com/ru/prompt-library/polyglot-superpowers): Переводите текст с любого языка на любой язык.
- [Поэт портманто](https://docs.anthropic.com/ru/prompt-library/portmanteau-poet): Объединение двух слов для создания нового, осмысленного портманто.
- [Профессионал по неймингу продуктов](https://docs.anthropic.com/ru/prompt-library/product-naming-pro): Создание привлекательных названий продуктов на основе описаний и ключевых слов
- [Шлифовщик текста](https://docs.anthropic.com/ru/prompt-library/prose-polisher): Улучшение и совершенствование письменного контента с помощью продвинутых методов редактирования и рекомендаций.
- [Pun-dit](https://docs.anthropic.com/ru/prompt-library/pun-dit): Генерация остроумных каламбуров и игры слов на любую заданную тему.
- [Отладчик Python](https://docs.anthropic.com/ru/prompt-library/python-bug-buster): Обнаружение и исправление ошибок в коде Python.
- [Классификатор отзывов](https://docs.anthropic.com/ru/prompt-library/review-classifier): Категоризация обратной связи по предопределенным тегам и категориям.
- [Загадай-ка мне загадку](https://docs.anthropic.com/ru/prompt-library/riddle-me-this): Создание загадок и помощь пользователю в поиске решений
- [Симулятор научно-фантастических сценариев](https://docs.anthropic.com/ru/prompt-library/sci-fi-scenario-simulator): Обсуждение с пользователем различных научно-фантастических сценариев и связанных с ними проблем и соображений.
- [Упроститель для второго класса](https://docs.anthropic.com/ru/prompt-library/second-grade-simplifier): Делает сложный текст понятным для юных учеников
- [Мастер сравнений](https://docs.anthropic.com/ru/prompt-library/simile-savant): Создание сравнений из базовых описаний
- [Сократовский мудрец](https://docs.anthropic.com/ru/prompt-library/socratic-sage): Ведение сократовской беседы на заданную пользователем тему
- [Волшебник электронных таблиц](https://docs.anthropic.com/ru/prompt-library/spreadsheet-sorcerer): Создание CSV-таблиц с различными типами данных.
- [SQL чародей](https://docs.anthropic.com/ru/prompt-library/sql-sorcerer): Преобразование повседневного языка в SQL-запросы
- [Помощник в рассказывании историй](https://docs.anthropic.com/ru/prompt-library/storytelling-sidekick): Совместное создание увлекательных историй с пользователем, предлагая неожиданные повороты сюжета и развитие персонажей
- [Консультант по путешествиям во времени](https://docs.anthropic.com/ru/prompt-library/time-travel-consultant): Помощь пользователю в навигации по гипотетическим сценариям путешествий во времени и их последствиям
- [Скороговорки](https://docs.anthropic.com/ru/prompt-library/tongue-twister): Создание сложных скороговорок
- [Генератор викторины](https://docs.anthropic.com/ru/prompt-library/trivia-generator): Создание вопросов для викторины на различные темы с подсказками при необходимости
- [Детектор тональности твитов](https://docs.anthropic.com/ru/prompt-library/tweet-tone-detector): Определение тона и настроения твитов.
- [Инноватор в сфере VR фитнеса](https://docs.anthropic.com/ru/prompt-library/vr-fitness-innovator): Генерация креативных идей для фитнес-игр в виртуальной реальности.
- [Мастер создания веб-сайтов](https://docs.anthropic.com/ru/prompt-library/website-wizard): Создание одностраничных веб-сайтов на основе пользовательских спецификаций.
- [API](https://docs.anthropic.com/ru/release-notes/api): Следите за обновлениями API Anthropic и Консоли разработчика.
- [Приложения Claude](https://docs.anthropic.com/ru/release-notes/claude-apps): Следите за обновлениями приложений Claude от Anthropic.
- [Обзор](https://docs.anthropic.com/ru/release-notes/overview): Следите за обновлениями продуктов и сервисов Anthropic.
- [Системные промпты](https://docs.anthropic.com/ru/release-notes/system-prompts): Смотрите обновления основных системных промптов на [Claude.ai](https://www.claude.ai) и в приложениях Claude для [iOS](http://anthropic.com/ios) и [Android](http://anthropic.com/android).
- [取消消息批次（测试版）](https://docs.anthropic.com/zh-CN/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/zh-CN/api/claude-on-amazon-bedrock): Anthropic的Claude模型现已通过Amazon Bedrock全面上市。
- [Vertex AI API](https://docs.anthropic.com/zh-CN/api/claude-on-vertex-ai): Anthropic的Claude模型现已通过[Vertex AI](https://cloud.google.com/vertex-ai)全面上线。
- [客户端 SDK](https://docs.anthropic.com/zh-CN/api/client-sdks): 我们提供 Python 和 TypeScript 库，使得使用 Anthropic API 变得更加容易。
- [Create a Text Completion](https://docs.anthropic.com/zh-CN/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [创建消息批次（测试版）](https://docs.anthropic.com/zh-CN/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [错误](https://docs.anthropic.com/zh-CN/api/errors)
- [获取帮助](https://docs.anthropic.com/zh-CN/api/getting-help): 我们已经在这些文档中尽量提供了最常见问题的答案。但是，如果您在使用Claude、Anthropic API或我们的任何产品时需要进一步的技术支持，您可以通过[support.anthropic.com](https://support.anthropic.com)联系我们的支持团队。
- [入门指南](https://docs.anthropic.com/zh-CN/api/getting-started)
- [IP地址](https://docs.anthropic.com/zh-CN/api/ip-addresses): Anthropic服务使用固定范围的IP地址。您可以将这些地址添加到防火墙中，以便在访问Anthropic API和控制台时开放最小的出口流量。这些范围在未经通知的情况下不会改变。
- [列出消息批次（测试版）](https://docs.anthropic.com/zh-CN/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/zh-CN/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [消息批次示例](https://docs.anthropic.com/zh-CN/api/messages-batch-examples): 消息批次API的使用示例
- [计算消息令牌数（测试版）](https://docs.anthropic.com/zh-CN/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [消息示例](https://docs.anthropic.com/zh-CN/api/messages-examples): 消息 API 的请求和响应示例
- [流式消息](https://docs.anthropic.com/zh-CN/api/messages-streaming)
- [从文本补全迁移](https://docs.anthropic.com/zh-CN/api/migrating-from-text-completions-to-messages): 从文本补全迁移到消息
- [提示验证](https://docs.anthropic.com/zh-CN/api/prompt-validation): 使用文本补全
- [使用限制](https://docs.anthropic.com/zh-CN/api/rate-limits): 为了减少滥用并管理我们API的容量，我们对组织使用Claude API的数量实施了限制。
- [获取消息批次结果（测试版）](https://docs.anthropic.com/zh-CN/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [获取消息批次（测试版）](https://docs.anthropic.com/zh-CN/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [流式文本补全](https://docs.anthropic.com/zh-CN/api/streaming)
- [支持的地区](https://docs.anthropic.com/zh-CN/api/supported-regions): 以下是我们目前支持访问的国家、地区和领土：
- [版本](https://docs.anthropic.com/zh-CN/api/versioning): 在发送API请求时，您必须发送一个`anthropic-version`请求头。例如，`anthropic-version: 2023-06-01`。如果您使用我们的[客户端库](/zh/api/client-libraries)，这将自动为您处理。
- [2024年7月更新](https://docs.anthropic.com/zh-CN/developer-newsletter/july2024): 2024年7月28日
- [概述](https://docs.anthropic.com/zh-CN/developer-newsletter/overview): 探索每月更新、工程深度解析、最佳实践和成功案例，以增强您的 Claude 集成
- [模型](https://docs.anthropic.com/zh-CN/docs/about-claude/models): Claude 是由 Anthropic 开发的一系列最先进的大型语言模型。本指南介绍我们的模型并将其性能与传统模型进行比较。
- [安全性和合规性](https://docs.anthropic.com/zh-CN/docs/about-claude/security-compliance)
- [内容审核](https://docs.anthropic.com/zh-CN/docs/about-claude/use-case-guides/content-moderation): 内容审核是在数字应用程序中维护安全、尊重和富有成效的环境的关键方面。在本指南中，我们将讨论如何使用 Claude 在您的数字应用程序中进行内容审核。
- [客服代理](https://docs.anthropic.com/zh-CN/docs/about-claude/use-case-guides/customer-support-chat): 本指南介绍如何利用Claude的高级对话能力实时处理客户咨询，提供全天候支持，减少等待时间，并通过准确的回应和积极的互动来处理大量支持请求。
- [法律文件总结](https://docs.anthropic.com/zh-CN/docs/about-claude/use-case-guides/legal-summarization): 本指南介绍如何利用Claude的高级自然语言处理能力来高效总结法律文件，提取关键信息并加快法律研究。通过Claude，您可以简化合同审查、诉讼准备和监管工作，节省时间并确保法律流程的准确性。
- [常见用例指南](https://docs.anthropic.com/zh-CN/docs/about-claude/use-case-guides/overview)
- [工单路由](https://docs.anthropic.com/zh-CN/docs/about-claude/use-case-guides/ticket-routing): 本指南介绍如何利用 Claude 的高级自然语言理解能力,根据客户意图、紧急程度、优先级、客户画像等因素大规模分类客服工单。
- [Google Sheets插件](https://docs.anthropic.com/zh-CN/docs/build-with-claude/claude-for-sheets): [Claude for Sheets扩展程序](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257)将Claude集成到Google Sheets中，让您可以直接在单元格中与Claude进行交互。
- [计算机使用 (测试版)](https://docs.anthropic.com/zh-CN/docs/build-with-claude/computer-use): 升级后的 Claude 3.5 Sonnet 模型能够与可操作计算机桌面环境的[工具](/zh/docs/build-with-claude/tool-use)进行交互。
- [定义你的成功标准](https://docs.anthropic.com/zh-CN/docs/build-with-claude/define-success)
- [创建有效的实证评估](https://docs.anthropic.com/zh-CN/docs/build-with-claude/develop-tests)
- [嵌入向量](https://docs.anthropic.com/zh-CN/docs/build-with-claude/embeddings): 文本嵌入是文本的数值表示,可用于衡量语义相似度。本指南介绍了嵌入向量、其应用场景,以及如何使用嵌入模型进行搜索、推荐和异常检测等任务。
- [消息批处理（测试版）](https://docs.anthropic.com/zh-CN/docs/build-with-claude/message-batches)
- [PDF支持（测试版）](https://docs.anthropic.com/zh-CN/docs/build-with-claude/pdf-support): Claude 3.5 Sonnet模型现在支持PDF输入，并能理解文档中的文本和视觉内容。
- [提示缓存（测试版）](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-caching)
- [保持清晰、直接和详细](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [让 Claude 思考（思维链提示）以提高性能](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [链式复杂提示以获得更强的性能](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/chain-prompts)
- [长文本提示技巧](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/long-context-tips)
- [使用示例（多示例提示）来引导Claude的行为](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [提示工程概述](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/overview)
- [预填充Claude的回复以获得更好的输出控制](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [自动生成初始提示词模板](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/prompt-generator)
- [使用我们的提示改进器来优化您的提示](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/prompt-improver)
- [使用提示模板和变量](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [通过系统提示给Claude分配角色](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/system-prompts)
- [使用XML标签构建你的提示](https://docs.anthropic.com/zh-CN/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [文本生成](https://docs.anthropic.com/zh-CN/docs/build-with-claude/text-generation): Claude在广泛的文本任务中表现出色。Claude经过训练可以处理代码、散文和其他自然语言输入，并提供相应的文本输出。
- [令牌计数（测试版）](https://docs.anthropic.com/zh-CN/docs/build-with-claude/token-counting)
- [工具使用（函数调用）](https://docs.anthropic.com/zh-CN/docs/build-with-claude/tool-use)
- [视觉功能](https://docs.anthropic.com/zh-CN/docs/build-with-claude/vision): Claude 3系列模型具有新的视觉功能，使Claude能够理解和分析图像，为多模态交互开启了令人兴奋的可能性。
- [初始设置](https://docs.anthropic.com/zh-CN/docs/initial-setup): 让我们学习如何使用 Anthropic API 来构建 Claude 应用
- [Claude简介](https://docs.anthropic.com/zh-CN/docs/intro-to-claude): Claude是由Anthropic开发的一系列[高性能且智能的AI模型](/zh/docs/about-claude/models)。虽然Claude功能强大且可扩展，但它也是目前最值得信赖和可靠的AI。它遵循关键协议，减少错误，并且能够抵御越狱攻击——使[企业客户](https://www.anthropic.com/customers)能够大规模构建最安全的AI驱动应用程序。
- [Anthropic 隐私政策](https://docs.anthropic.com/zh-CN/docs/legal-center/privacy)
- [Anthropic 指南](https://docs.anthropic.com/zh-CN/docs/resources/cookbook)
- [Anthropic 课程](https://docs.anthropic.com/zh-CN/docs/resources/courses)
- [术语表](https://docs.anthropic.com/zh-CN/docs/resources/glossary): 这些概念并非 Anthropic 语言模型所独有，但我们在下面简要总结了关键术语。
- [Claude 3 模型说明卡](https://docs.anthropic.com/zh-CN/docs/resources/model-card)
- [模型弃用](https://docs.anthropic.com/zh-CN/docs/resources/model-deprecations)
- [系统状态](https://docs.anthropic.com/zh-CN/docs/resources/status)
- [使用评估工具](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/eval-tool): [Anthropic Console](https://console.anthropic.com/dashboard)提供了一个**评估工具**，让您可以在各种场景下测试您的提示。
- [提高输出一致性（JSON模式）](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [通过角色提示和预填充保持Claude的角色特征](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [防范越狱和提示注入](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [减少幻觉](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [降低延迟](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [减少提示词泄露](https://docs.anthropic.com/zh-CN/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [欢迎使用 Claude](https://docs.anthropic.com/zh-CN/docs/welcome): Claude 是由 Anthropic 打造的高性能、可信赖且智能的 AI 平台。Claude 在语言、推理、分析、编程等任务方面表现出色。
- [null](https://docs.anthropic.com/zh-CN/home)
- [自适应编辑器](https://docs.anthropic.com/zh-CN/prompt-library/adaptive-editor): 根据用户给出的指示重写文本，例如使用不同的语气、受众或风格。
- [机场代码分析器](https://docs.anthropic.com/zh-CN/prompt-library/airport-code-analyst): 从文本中查找和提取机场代码。
- [外星人类学家](https://docs.anthropic.com/zh-CN/prompt-library/alien-anthropologist): 从外星人类学家的视角分析人类文化和习俗。
- [头韵炼金术师](https://docs.anthropic.com/zh-CN/prompt-library/alliteration-alchemist): 为任何给定主题生成头韵短语和句子。
- [Babel的广播](https://docs.anthropic.com/zh-CN/prompt-library/babels-broadcasts): 用世界十大最常用语言创建引人注目的产品发布推文。
- [品牌构建器](https://docs.anthropic.com/zh-CN/prompt-library/brand-builder): 为整体品牌标识制定设计方案。
- [职业教练](https://docs.anthropic.com/zh-CN/prompt-library/career-coach): 与AI职业教练进行角色扮演对话
- [引用你的来源](https://docs.anthropic.com/zh-CN/prompt-library/cite-your-sources): 获取关于文档内容的问题答案，并提供支持回答的相关引用。
- [代码解释器](https://docs.anthropic.com/zh-CN/prompt-library/code-clarifier): 用简单的语言解释和简化复杂的代码。
- [代码顾问](https://docs.anthropic.com/zh-CN/prompt-library/code-consultant): 建议改进以优化Python代码性能。
- [企业洞察者](https://docs.anthropic.com/zh-CN/prompt-library/corporate-clairvoyant): 从冗长的企业报告中提取见解、识别风险并将关键信息提炼为单一备忘录。
- [宇宙按键](https://docs.anthropic.com/zh-CN/prompt-library/cosmic-keystrokes): 生成一个交互式打字速度游戏,使用单个HTML文件,具有横向滚动游戏玩法和Tailwind CSS样式。
- [CSV转换器](https://docs.anthropic.com/zh-CN/prompt-library/csv-converter): 将各种格式(JSON、XML等)的数据转换为格式正确的CSV文件。
- [美食创作者](https://docs.anthropic.com/zh-CN/prompt-library/culinary-creator): 根据用户可用的食材和饮食偏好推荐食谱创意。
- [数据整理器](https://docs.anthropic.com/zh-CN/prompt-library/data-organizer): 将非结构化文本转换为定制的JSON表格。
- [方向解码器](https://docs.anthropic.com/zh-CN/prompt-library/direction-decoder): 将自然语言转换为逐步说明。
- [解梦专家](https://docs.anthropic.com/zh-CN/prompt-library/dream-interpreter): 为用户的梦境提供解释和象征意义的见解
- [效率估算器](https://docs.anthropic.com/zh-CN/prompt-library/efficiency-estimator): 计算函数和算法的时间复杂度。
- [电子邮件提取器](https://docs.anthropic.com/zh-CN/prompt-library/email-extractor): 从文档中提取电子邮件地址并转换为JSON格式列表。
- [表情符号编码器](https://docs.anthropic.com/zh-CN/prompt-library/emoji-encoder): 将纯文本转换为有趣且富有表现力的表情符号消息。
- [伦理困境导航器](https://docs.anthropic.com/zh-CN/prompt-library/ethical-dilemma-navigator): 帮助用户思考复杂的伦理困境并提供不同的视角
- [Excel公式专家](https://docs.anthropic.com/zh-CN/prompt-library/excel-formula-expert): 根据用户描述的计算或数据操作创建Excel公式。
- [函数生成器](https://docs.anthropic.com/zh-CN/prompt-library/function-fabricator): 根据详细规范创建Python函数。
- [未来主义时尚顾问](https://docs.anthropic.com/zh-CN/prompt-library/futuristic-fashion-advisor): 根据用户的具体偏好推荐前卫时尚潮流和风格
- [精通Git](https://docs.anthropic.com/zh-CN/prompt-library/git-gud): 根据用户描述的版本控制操作生成适当的Git命令。
- [Google Apps 脚本编写器](https://docs.anthropic.com/zh-CN/prompt-library/google-apps-scripter): 根据用户需求生成 Google Apps 脚本来完成任务。
- [评分专家](https://docs.anthropic.com/zh-CN/prompt-library/grading-guru): 根据用户定义的标准和要求比较和评估文本质量
- [语法精灵](https://docs.anthropic.com/zh-CN/prompt-library/grammar-genie): 将语法不正确的句子转换为正确的英语。
- [幽默助手Hal](https://docs.anthropic.com/zh-CN/prompt-library/hal-the-humorous-helper): 与一个知识渊博且带有讽刺幽默感的AI聊天。
- [成语解释器](https://docs.anthropic.com/zh-CN/prompt-library/idiom-illuminator): 解释常见成语和谚语的含义及来源
- [面试问题生成器](https://docs.anthropic.com/zh-CN/prompt-library/interview-question-crafter): 生成面试问题
- [LaTeX 图例](https://docs.anthropic.com/zh-CN/prompt-library/latex-legend): 编写 LaTeX 文档，生成数学方程式、表格等代码。
- [课程计划制定器](https://docs.anthropic.com/zh-CN/prompt-library/lesson-planner): 制定任何学科的深入课程计划
- [提示库](https://docs.anthropic.com/zh-CN/prompt-library/library)
- [主审仲裁者](https://docs.anthropic.com/zh-CN/prompt-library/master-moderator): 评估用户输入是否包含潜在有害或非法内容
- [会议记录员](https://docs.anthropic.com/zh-CN/prompt-library/meeting-scribe): 将会议内容提炼为简明摘要，包括讨论主题、关键要点和行动项目。
- [备忘录大师](https://docs.anthropic.com/zh-CN/prompt-library/memo-maestro): 根据关键要点撰写全面的公司备忘录。
- [正念导师](https://docs.anthropic.com/zh-CN/prompt-library/mindfulness-mentor): 指导用户进行正念练习和减压技巧
- [情绪色彩转换器](https://docs.anthropic.com/zh-CN/prompt-library/mood-colorizer): 将情绪文字描述转换为相应的十六进制颜色代码。
- [激励缪斯](https://docs.anthropic.com/zh-CN/prompt-library/motivational-muse): 根据用户输入提供个性化的激励信息和肯定语
- [新词创造器](https://docs.anthropic.com/zh-CN/prompt-library/neologism-creator): 基于用户提供的概念或想法发明新词并提供其定义。
- [观点权衡器](https://docs.anthropic.com/zh-CN/prompt-library/perspectives-ponderer): 权衡用户提供主题的利弊
- [哲学思考](https://docs.anthropic.com/zh-CN/prompt-library/philosophical-musings): 进行深入的哲学讨论和思想实验
- [PII 净化器](https://docs.anthropic.com/zh-CN/prompt-library/pii-purifier): 自动检测并删除文本文档中的个人身份信息(PII)
- [多语言超能力](https://docs.anthropic.com/zh-CN/prompt-library/polyglot-superpowers): 将任何语言的文本翻译成任何语言。
- [组合词诗人](https://docs.anthropic.com/zh-CN/prompt-library/portmanteau-poet): 将两个词语组合在一起创造一个新的、有意义的组合词。
- [产品命名专家](https://docs.anthropic.com/zh-CN/prompt-library/product-naming-pro): 从描述和关键词创建朗朗上口的产品名称
- [文章润色器](https://docs.anthropic.com/zh-CN/prompt-library/prose-polisher): 使用高级文字编辑技巧和建议来完善和改进书面内容。
- [双关语大师](https://docs.anthropic.com/zh-CN/prompt-library/pun-dit): 基于任何给定主题生成巧妙的双关语和文字游戏。
- [Python 代码调试器](https://docs.anthropic.com/zh-CN/prompt-library/python-bug-buster): 检测并修复 Python 代码中的错误。
- [评论分类器](https://docs.anthropic.com/zh-CN/prompt-library/review-classifier): 将反馈分类为预设标签和类别。
- [猜谜语](https://docs.anthropic.com/zh-CN/prompt-library/riddle-me-this): 生成谜语并引导用户找到答案
- [科幻场景模拟器](https://docs.anthropic.com/zh-CN/prompt-library/sci-fi-scenario-simulator): 与用户讨论各种科幻场景及其相关挑战和考虑因素。
- [二年级简化器](https://docs.anthropic.com/zh-CN/prompt-library/second-grade-simplifier): 让复杂文本变得容易被年轻学习者理解
- [比喻大师](https://docs.anthropic.com/zh-CN/prompt-library/simile-savant): 从基本描述生成比喻句
- [苏格拉底式智者](https://docs.anthropic.com/zh-CN/prompt-library/socratic-sage): 就用户给定的主题进行苏格拉底式对话
- [电子表格魔法师](https://docs.anthropic.com/zh-CN/prompt-library/spreadsheet-sorcerer): 生成包含各种类型数据的CSV电子表格。
- [SQL魔法师](https://docs.anthropic.com/zh-CN/prompt-library/sql-sorcerer): 将日常语言转换为SQL查询。
- [故事创作助手](https://docs.anthropic.com/zh-CN/prompt-library/storytelling-sidekick): 与用户协作创作引人入胜的故事，提供情节转折和人物发展建议。
- [时间旅行顾问](https://docs.anthropic.com/zh-CN/prompt-library/time-travel-consultant): 帮助用户探索假设性时间旅行场景及其影响。
- [绕口令](https://docs.anthropic.com/zh-CN/prompt-library/tongue-twister): 创建富有挑战性的绕口令。
- [趣味问答生成器](https://docs.anthropic.com/zh-CN/prompt-library/trivia-generator): 生成各种主题的趣味问答题目并在需要时提供提示。
- [推文语气检测器](https://docs.anthropic.com/zh-CN/prompt-library/tweet-tone-detector): 检测推文背后的语气和情感。
- [VR健身创新者](https://docs.anthropic.com/zh-CN/prompt-library/vr-fitness-innovator): 为虚拟现实健身游戏集思广益创意点子。
- [网站向导](https://docs.anthropic.com/zh-CN/prompt-library/website-wizard): 根据用户规格创建单页网站。
- [API](https://docs.anthropic.com/zh-CN/release-notes/api): 关注Anthropic的API和开发者控制台的最新更新。
- [Claude 应用](https://docs.anthropic.com/zh-CN/release-notes/claude-apps): 关注 Anthropic 的 Claude 应用的最新更新。
- [概述](https://docs.anthropic.com/zh-CN/release-notes/overview): 关注Anthropic产品和服务的最新动态。
- [系统提示](https://docs.anthropic.com/zh-CN/release-notes/system-prompts): 查看[Claude.ai](https://www.claude.ai)和Claude [iOS](http://anthropic.com/ios)及[Android](http://anthropic.com/android)应用程序的核心系统提示更新。
- [取消訊息批次（測試版）](https://docs.anthropic.com/zh-TW/api/canceling-message-batches): Batches may be canceled any time before processing ends. Once cancellation is initiated, the batch enters a `canceling` state, at which time the system may complete any in-progress, non-interruptible requests before finalizing cancellation.

The number of canceled requests is specified in `request_counts`. To determine which requests were canceled, check the individual results within the batch. Note that cancellation may not result in any canceled requests if they were non-interruptible.
- [Amazon Bedrock API](https://docs.anthropic.com/zh-TW/api/claude-on-amazon-bedrock): Anthropic 的 Claude 模型現已通過 Amazon Bedrock 全面開放使用。
- [Vertex AI API](https://docs.anthropic.com/zh-TW/api/claude-on-vertex-ai): Anthropic 的 Claude 模型現已通過 [Vertex AI](https://cloud.google.com/vertex-ai) 全面提供。
- [客戶端 SDK](https://docs.anthropic.com/zh-TW/api/client-sdks): 我們提供 Python 和 TypeScript 程式庫，讓您更容易使用 Anthropic API。
- [Create a Text Completion](https://docs.anthropic.com/zh-TW/api/complete): [Legacy] Create a Text Completion.

The Text Completions API is a legacy API. We recommend using the [Messages API](https://docs.anthropic.com/en/api/messages) going forward.

Future models and features will not be compatible with Text Completions. See our [migration guide](https://docs.anthropic.com/en/api/migrating-from-text-completions-to-messages) for guidance in migrating from Text Completions to Messages.
- [建立訊息批次（測試版）](https://docs.anthropic.com/zh-TW/api/creating-message-batches): Send a batch of Message creation requests.

The Message Batches API can be used to process multiple Messages API requests at once. Once a Message Batch is created, it begins processing immediately. Batches can take up to 24 hours to complete.
- [錯誤](https://docs.anthropic.com/zh-TW/api/errors)
- [取得協助](https://docs.anthropic.com/zh-TW/api/getting-help): 我們已在這些文件中嘗試提供最常見問題的解答。然而，如果您在使用 Claude、Anthropic API 或我們的任何產品時需要進一步的技術支援，您可以透過 [support.anthropic.com](https://support.anthropic.com) 聯繫我們的支援團隊。
- [開始使用](https://docs.anthropic.com/zh-TW/api/getting-started)
- [IP位址](https://docs.anthropic.com/zh-TW/api/ip-addresses): Anthropic服務使用固定範圍的IP位址。您可以將這些位址添加到防火牆中，以便在訪問Anthropic API和控制台時開放最小的出站流量範圍。這些範圍在未經通知的情況下不會改變。
- [列出訊息批次（測試版）](https://docs.anthropic.com/zh-TW/api/listing-message-batches): List all Message Batches within a Workspace. Most recently created batches are returned first.
- [Create a Message](https://docs.anthropic.com/zh-TW/api/messages): Send a structured list of input messages with text and/or image content, and the model will generate the next message in the conversation.

The Messages API can be used for either single queries or stateless multi-turn conversations.
- [訊息批次範例](https://docs.anthropic.com/zh-TW/api/messages-batch-examples): 訊息批次 API 的使用範例
- [計算訊息字符數（測試版）](https://docs.anthropic.com/zh-TW/api/messages-count-tokens): Count the number of tokens in a Message.

The Token Count API can be used to count the number of tokens in a Message, including tools, images, and documents, without creating it.
- [訊息範例](https://docs.anthropic.com/zh-TW/api/messages-examples): 訊息 API 的請求和回應範例
- [串流訊息](https://docs.anthropic.com/zh-TW/api/messages-streaming)
- [從文本補全遷移](https://docs.anthropic.com/zh-TW/api/migrating-from-text-completions-to-messages): 從文本補全遷移到訊息
- [提示詞驗證](https://docs.anthropic.com/zh-TW/api/prompt-validation): 使用文本完成功能
- [使用限制](https://docs.anthropic.com/zh-TW/api/rate-limits): 為了減少濫用並管理我們API的容量，我們對組織使用Claude API的程度設置了限制。
- [檢索訊息批次結果（測試版）](https://docs.anthropic.com/zh-TW/api/retrieving-message-batch-results): Streams the results of a Message Batch as a `.jsonl` file.

Each line in the file is a JSON object containing the result of a single request in the Message Batch. Results are not guaranteed to be in the same order as requests. Use the `custom_id` field to match results to requests.
- [檢索訊息批次（測試版）](https://docs.anthropic.com/zh-TW/api/retrieving-message-batches): This endpoint is idempotent and can be used to poll for Message Batch completion. To access the results of a Message Batch, make a request to the `results_url` field in the response.
- [串流文字補全](https://docs.anthropic.com/zh-TW/api/streaming)
- [支援地區](https://docs.anthropic.com/zh-TW/api/supported-regions): 以下是我們目前可以支援存取的國家、地區和領土：
- [版本](https://docs.anthropic.com/zh-TW/api/versioning): 在發送 API 請求時，您必須發送一個 `anthropic-version` 請求標頭。例如，`anthropic-version: 2023-06-01`。如果您使用我們的[客戶端函式庫](/zh/api/client-libraries)，這將自動為您處理。
- [2024年7月更新](https://docs.anthropic.com/zh-TW/developer-newsletter/july2024): 2024年7月28日
- [概述](https://docs.anthropic.com/zh-TW/developer-newsletter/overview): 探索每月更新、工程深入探討、最佳實踐和成功案例，以增強您的 Claude 整合
- [模型](https://docs.anthropic.com/zh-TW/docs/about-claude/models): Claude 是由 Anthropic 開發的一系列最先進的大型語言模型。本指南介紹我們的模型並將其性能與舊版模型進行比較。
- [安全性與合規性](https://docs.anthropic.com/zh-TW/docs/about-claude/security-compliance)
- [內容審核](https://docs.anthropic.com/zh-TW/docs/about-claude/use-case-guides/content-moderation): 內容審核是在數位應用程式中維護安全、尊重和高效環境的關鍵要素。在本指南中，我們將討論如何使用 Claude 在您的數位應用程式中進行內容審核。
- [客戶支援代理](https://docs.anthropic.com/zh-TW/docs/about-claude/use-case-guides/customer-support-chat): 本指南說明如何利用 Claude 的進階對話功能即時處理客戶查詢，提供全天候支援，縮短等待時間，並透過準確的回應和正面的互動來管理大量支援需求。
- [法律文件摘要](https://docs.anthropic.com/zh-TW/docs/about-claude/use-case-guides/legal-summarization): 本指南說明如何利用 Claude 的進階自然語言處理能力來高效地總結法律文件、提取關鍵資訊並加快法律研究。透過 Claude，您可以簡化合約審查、訴訟準備和法規工作，節省時間並確保法律流程的準確性。
- [常見使用案例指南](https://docs.anthropic.com/zh-TW/docs/about-claude/use-case-guides/overview)
- [工單路由](https://docs.anthropic.com/zh-TW/docs/about-claude/use-case-guides/ticket-routing): 本指南說明如何利用 Claude 的進階自然語言理解能力，根據客戶意圖、緊急程度、優先順序、客戶資料等因素大規模分類客戶支援工單。
- [Google Sheets 附加元件](https://docs.anthropic.com/zh-TW/docs/build-with-claude/claude-for-sheets): [Claude for Sheets 擴充功能](https://workspace.google.com/marketplace/app/claude%5Ffor%5Fsheets/909417792257)將 Claude 整合到 Google Sheets 中，讓您可以直接在儲存格中執行與 Claude 的互動。
- [電腦使用 (beta)](https://docs.anthropic.com/zh-TW/docs/build-with-claude/computer-use): 升級後的 Claude 3.5 Sonnet 模型能夠與可操作電腦桌面環境的[工具](/zh-tw/docs/build-with-claude/tool-use)互動。
- [定義你的成功標準](https://docs.anthropic.com/zh-TW/docs/build-with-claude/define-success)
- [建立強大的實證評估](https://docs.anthropic.com/zh-TW/docs/build-with-claude/develop-tests)
- [嵌入向量](https://docs.anthropic.com/zh-TW/docs/build-with-claude/embeddings): 文本嵌入是文本的數值表示，可用於衡量語義相似度。本指南介紹嵌入向量、其應用場景，以及如何使用嵌入模型進行搜索、推薦和異常檢測等任務。
- [訊息批次處理 (beta)](https://docs.anthropic.com/zh-TW/docs/build-with-claude/message-batches)
- [PDF 支援功能（測試版）](https://docs.anthropic.com/zh-TW/docs/build-with-claude/pdf-support): Claude 3.5 Sonnet 模型現在支援 PDF 輸入，並能理解文件中的文字和視覺內容。
- [提示詞快取（測試版）](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-caching)
- [保持清晰、直接和詳細](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/be-clear-and-direct)
- [讓 Claude 思考（思維鏈提示）以提升表現](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/chain-of-thought)
- [串聯複雜提示以獲得更強大的表現](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/chain-prompts)
- [長文本提示詞使用技巧](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/long-context-tips)
- [使用範例（多樣提示）來引導 Claude 的行為](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/multishot-prompting)
- [提示工程概述](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/overview)
- [預填 Claude 的回應以獲得更好的輸出控制](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/prefill-claudes-response)
- [自動生成初稿提示詞模板](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/prompt-generator)
- [使用我們的提示改進工具來優化您的提示](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/prompt-improver)
- [使用提示詞模板和變數](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/prompt-templates-and-variables)
- [透過系統提示賦予 Claude 角色](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/system-prompts)
- [使用XML標籤來結構化您的提示](https://docs.anthropic.com/zh-TW/docs/build-with-claude/prompt-engineering/use-xml-tags)
- [文字生成](https://docs.anthropic.com/zh-TW/docs/build-with-claude/text-generation): Claude 在各種文字相關任務中表現出色。Claude 經過訓練，能夠處理程式碼、散文和其他自然語言輸入，並提供相應的文字輸出。
- [代幣計數（測試版）](https://docs.anthropic.com/zh-TW/docs/build-with-claude/token-counting)
- [工具使用（函數呼叫）](https://docs.anthropic.com/zh-TW/docs/build-with-claude/tool-use)
- [視覺能力](https://docs.anthropic.com/zh-TW/docs/build-with-claude/vision): Claude 3系列模型具備新的視覺能力，使Claude能夠理解和分析圖像，為多模態互動開創令人興奮的可能性。
- [初始設定](https://docs.anthropic.com/zh-TW/docs/initial-setup): 讓我們學習如何使用 Anthropic API 來建構 Claude
- [Claude 簡介](https://docs.anthropic.com/zh-TW/docs/intro-to-claude): Claude 是由 Anthropic 開發的一系列[高性能且智能的 AI 模型](/zh-tw/docs/about-claude/models)。雖然 Claude 功能強大且可擴展，但它也是目前最值得信賴和可靠的 AI。它遵循關鍵協議，較少出錯，且能抵禦越獄攻擊—讓[企業客戶](https://www.anthropic.com/customers)能夠大規模建立最安全的 AI 驅動應用程式。
- [Anthropic 隱私權政策](https://docs.anthropic.com/zh-TW/docs/legal-center/privacy)
- [Anthropic 食譜](https://docs.anthropic.com/zh-TW/docs/resources/cookbook)
- [Anthropic 課程](https://docs.anthropic.com/zh-TW/docs/resources/courses)
- [詞彙表](https://docs.anthropic.com/zh-TW/docs/resources/glossary): 這些概念並非 Anthropic 語言模型獨有，但我們在下面簡要總結了關鍵術語。
- [Claude 3 模型說明卡](https://docs.anthropic.com/zh-TW/docs/resources/model-card)
- [模型淘汰](https://docs.anthropic.com/zh-TW/docs/resources/model-deprecations)
- [系統狀態](https://docs.anthropic.com/zh-TW/docs/resources/status)
- [使用評估工具](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/eval-tool): [Anthropic Console](https://console.anthropic.com/dashboard) 提供了一個**評估工具**，讓您可以在各種情境下測試您的提示詞。
- [提高輸出一致性（JSON模式）](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/strengthen-guardrails/increase-consistency)
- [透過角色提示和預填來保持 Claude 的角色特性](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/strengthen-guardrails/keep-claude-in-character)
- [防範越獄和提示詞注入](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/strengthen-guardrails/mitigate-jailbreaks)
- [減少幻覺](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/strengthen-guardrails/reduce-hallucinations)
- [降低延遲](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/strengthen-guardrails/reduce-latency)
- [減少提示詞洩漏](https://docs.anthropic.com/zh-TW/docs/test-and-evaluate/strengthen-guardrails/reduce-prompt-leak)
- [歡迎使用 Claude](https://docs.anthropic.com/zh-TW/docs/welcome): Claude 是由 Anthropic 打造的高性能、可信賴且智能的 AI 平台。Claude 在語言、推理、分析、編程等任務方面表現出色。
- [null](https://docs.anthropic.com/zh-TW/home)
- [自適應編輯器](https://docs.anthropic.com/zh-TW/prompt-library/adaptive-editor): 根據用戶給定的指示重寫文本，例如使用不同的語氣、受眾或風格。
- [機場代碼分析器](https://docs.anthropic.com/zh-TW/prompt-library/airport-code-analyst): 從文本中查找和提取機場代碼。
- [外星人類學家](https://docs.anthropic.com/zh-TW/prompt-library/alien-anthropologist): 從外星人類學家的角度分析人類文化和習俗。
- [頭韻詞煉金術師](https://docs.anthropic.com/zh-TW/prompt-library/alliteration-alchemist): 為任何給定主題生成頭韻短語和句子。
- [巴別塔的廣播](https://docs.anthropic.com/zh-TW/prompt-library/babels-broadcasts): 用世界十大最常用語言創建引人注目的產品發布推文
- [品牌建構器](https://docs.anthropic.com/zh-TW/prompt-library/brand-builder): 為全方位品牌識別設計製作設計簡報。
- [職業生涯教練](https://docs.anthropic.com/zh-TW/prompt-library/career-coach): 與AI職業生涯教練進行角色扮演對話
- [引用你的來源](https://docs.anthropic.com/zh-TW/prompt-library/cite-your-sources): 獲取有關文件內容的問題答案，並附上支持回應的相關引用。
- [程式碼解釋器](https://docs.anthropic.com/zh-TW/prompt-library/code-clarifier): 以簡單的語言解釋和簡化複雜的程式碼。
- [程式碼顧問](https://docs.anthropic.com/zh-TW/prompt-library/code-consultant): 建議改進以優化 Python 程式碼效能
- [企業洞察者](https://docs.anthropic.com/zh-TW/prompt-library/corporate-clairvoyant): 從冗長的企業報告中提取見解、識別風險，並將關鍵信息濃縮為單一備忘錄。
- [宇宙按鍵](https://docs.anthropic.com/zh-TW/prompt-library/cosmic-keystrokes): 在單一HTML檔案中生成一個互動式打字速度遊戲，具有側捲式遊戲玩法和Tailwind CSS樣式。
- [CSV 轉換器](https://docs.anthropic.com/zh-TW/prompt-library/csv-converter): 將各種格式（JSON、XML等）的數據轉換為正確格式的 CSV 文件。
- [美食創作者](https://docs.anthropic.com/zh-TW/prompt-library/culinary-creator): 根據用戶可用的食材和飲食偏好提供食譜建議。
- [資料整理器](https://docs.anthropic.com/zh-TW/prompt-library/data-organizer): 將非結構化文字轉換為客製化 JSON 表格。
- [方向解碼器](https://docs.anthropic.com/zh-TW/prompt-library/direction-decoder): 將自然語言轉換為逐步說明。
- [解夢者](https://docs.anthropic.com/zh-TW/prompt-library/dream-interpreter): 為用戶的夢境提供解釋和象徵意義的見解
- [效率估算器](https://docs.anthropic.com/zh-TW/prompt-library/efficiency-estimator): 計算函數和演算法的時間複雜度。
- [電子郵件提取器](https://docs.anthropic.com/zh-TW/prompt-library/email-extractor): 從文件中提取電子郵件地址並轉換為JSON格式列表。
- [表情符號編碼器](https://docs.anthropic.com/zh-TW/prompt-library/emoji-encoder): 將純文字轉換成有趣且富有表現力的表情符號訊息。
- [倫理困境導航器](https://docs.anthropic.com/zh-TW/prompt-library/ethical-dilemma-navigator): 幫助用戶思考複雜的倫理困境並提供不同的觀點
- [Excel公式專家](https://docs.anthropic.com/zh-TW/prompt-library/excel-formula-expert): 根據用戶描述的計算或數據操作創建Excel公式。
- [函數生成器](https://docs.anthropic.com/zh-TW/prompt-library/function-fabricator): 根據詳細規格創建 Python 函數。
- [未來主義時尚顧問](https://docs.anthropic.com/zh-TW/prompt-library/futuristic-fashion-advisor): 根據用戶的具體偏好建議前衛時尚趨勢和風格
- [精通 Git](https://docs.anthropic.com/zh-TW/prompt-library/git-gud): 根據使用者描述的版本控制操作生成適當的 Git 命令。
- [Google Apps 腳本編寫器](https://docs.anthropic.com/zh-TW/prompt-library/google-apps-scripter): 根據用戶需求生成 Google Apps 腳本以完成任務。
- [評分專家](https://docs.anthropic.com/zh-TW/prompt-library/grading-guru): 根據用戶定義的標準和準則比較和評估文本的質量
- [文法精靈](https://docs.anthropic.com/zh-TW/prompt-library/grammar-genie): 將語法不正確的句子轉換為正確的英語。
- [幽默助手Hal](https://docs.anthropic.com/zh-TW/prompt-library/hal-the-humorous-helper): 與一位具有諷刺性格的知識豐富AI聊天。
- [成語解析器](https://docs.anthropic.com/zh-TW/prompt-library/idiom-illuminator): 解釋常見成語和諺語的含義及來源
- [面試問題生成器](https://docs.anthropic.com/zh-TW/prompt-library/interview-question-crafter): 生成面試問題。
- [LaTeX 圖例](https://docs.anthropic.com/zh-TW/prompt-library/latex-legend): 撰寫 LaTeX 文件，生成數學方程式、表格等代碼。
- [課程計劃制定工具](https://docs.anthropic.com/zh-TW/prompt-library/lesson-planner): 在任何學科上制定深入的課程計劃。
- [提示庫](https://docs.anthropic.com/zh-TW/prompt-library/library)
- [主要審核者](https://docs.anthropic.com/zh-TW/prompt-library/master-moderator): 評估用戶輸入是否包含潛在有害或非法內容
- [會議記錄員](https://docs.anthropic.com/zh-TW/prompt-library/meeting-scribe): 將會議內容提煉為簡潔的摘要，包括討論主題、重要結論和行動項目。
- [備忘錄大師](https://docs.anthropic.com/zh-TW/prompt-library/memo-maestro): 根據關鍵要點撰寫全面的公司備忘錄
- [正念導師](https://docs.anthropic.com/zh-TW/prompt-library/mindfulness-mentor): 引導用戶進行正念練習和減壓技巧
- [情緒色彩轉換器](https://docs.anthropic.com/zh-TW/prompt-library/mood-colorizer): 將情緒文字描述轉換為相應的十六進制顏色代碼。
- [激勵繆思](https://docs.anthropic.com/zh-TW/prompt-library/motivational-muse): 根據用戶輸入提供個性化的激勵訊息和肯定語
- [新詞創造器](https://docs.anthropic.com/zh-TW/prompt-library/neologism-creator): 根據用戶提供的概念或想法發明新詞並提供其定義。
- [觀點權衡者](https://docs.anthropic.com/zh-TW/prompt-library/perspectives-ponderer): 權衡使用者提供主題的利弊。
- [哲學思考](https://docs.anthropic.com/zh-TW/prompt-library/philosophical-musings): 進行深度哲學討論和思想實驗。
- [PII 淨化器](https://docs.anthropic.com/zh-TW/prompt-library/pii-purifier): 自動檢測並移除文本文件中的個人身份識別資訊 (PII)。
- [多語言超能力](https://docs.anthropic.com/zh-TW/prompt-library/polyglot-superpowers): 將任何語言翻譯成任何語言。
- [組合詞詩人](https://docs.anthropic.com/zh-TW/prompt-library/portmanteau-poet): 將兩個詞組合在一起創造一個新的、有意義的組合詞。
- [產品命名專家](https://docs.anthropic.com/zh-TW/prompt-library/product-naming-pro): 從描述和關鍵字創建引人注目的產品名稱
- [文章修飾器](https://docs.anthropic.com/zh-TW/prompt-library/prose-polisher): 使用進階文字編輯技巧和建議來改進和提升書面內容。
- [雙關語大師](https://docs.anthropic.com/zh-TW/prompt-library/pun-dit): 根據任何給定主題生成巧妙的雙關語和文字遊戲。
- [Python 除錯工具](https://docs.anthropic.com/zh-TW/prompt-library/python-bug-buster): 檢測並修復 Python 程式碼中的錯誤。
- [評論分類器](https://docs.anthropic.com/zh-TW/prompt-library/review-classifier): 將反饋分類為預先指定的標籤和類別。
- [來猜個謎語吧](https://docs.anthropic.com/zh-TW/prompt-library/riddle-me-this): 生成謎語並引導使用者找到答案
- [科幻場景模擬器](https://docs.anthropic.com/zh-TW/prompt-library/sci-fi-scenario-simulator): 與用戶討論各種科幻場景及其相關挑戰和考慮因素。
- [二年級簡化器](https://docs.anthropic.com/zh-TW/prompt-library/second-grade-simplifier): 讓複雜的文字變得容易讓年輕學習者理解
- [比喻大師](https://docs.anthropic.com/zh-TW/prompt-library/simile-savant): 從基本描述生成比喻。
- [蘇格拉底式智者](https://docs.anthropic.com/zh-TW/prompt-library/socratic-sage): 以蘇格拉底式對話方式討論用戶給定的主題
- [試算表魔術師](https://docs.anthropic.com/zh-TW/prompt-library/spreadsheet-sorcerer): 生成包含各種類型數據的 CSV 試算表。
- [SQL 魔法師](https://docs.anthropic.com/zh-TW/prompt-library/sql-sorcerer): 將日常語言轉換為 SQL 查詢。
- [說故事小幫手](https://docs.anthropic.com/zh-TW/prompt-library/storytelling-sidekick): 與使用者協作創作引人入勝的故事，提供情節轉折和角色發展。
- [時間旅行顧問](https://docs.anthropic.com/zh-TW/prompt-library/time-travel-consultant): 幫助用戶探索假設性時間旅行場景及其影響。
- [繞口令](https://docs.anthropic.com/zh-TW/prompt-library/tongue-twister): 創建富有挑戰性的繞口令。
- [問答遊戲生成器](https://docs.anthropic.com/zh-TW/prompt-library/trivia-generator): 在各種主題上生成問答題目，並在需要時提供提示。
- [推文語氣偵測器](https://docs.anthropic.com/zh-TW/prompt-library/tweet-tone-detector): 偵測推文背後的語氣和情感。
- [VR 健身創新者](https://docs.anthropic.com/zh-TW/prompt-library/vr-fitness-innovator): 為虛擬實境健身遊戲集思廣益創意點子。
- [網站精靈](https://docs.anthropic.com/zh-TW/prompt-library/website-wizard): 根據用戶規格創建單頁網站。
- [API](https://docs.anthropic.com/zh-TW/release-notes/api): 追蹤 Anthropic 的 API 和開發者控制台的最新更新。
- [Claude 應用程式](https://docs.anthropic.com/zh-TW/release-notes/claude-apps): 追蹤 Anthropic 的 Claude 應用程式的最新更新。
- [概述](https://docs.anthropic.com/zh-TW/release-notes/overview): 追蹤 Anthropic 產品和服務的最新更新。
- [系統提示詞](https://docs.anthropic.com/zh-TW/release-notes/system-prompts): 查看 [Claude.ai](https://www.claude.ai) 和 Claude [iOS](http://anthropic.com/ios) 及 [Android](http://anthropic.com/android) 應用程式的核心系統提示詞更新。


## Optional

- [Developer Console](https://console.anthropic.com/)
- [Developer Discord](https://www.anthropic.com/discord)
- [Support](https://support.anthropic.com/)
- [Console Développeur](https://console.anthropic.com/)
- [Discord pour les développeurs](https://www.anthropic.com/discord)
- [Assistance](https://support.anthropic.com/)
- [Entwicklerkonsole](https://console.anthropic.com/)
- [Entwickler-Discord](https://www.anthropic.com/discord)
- [Unterstützung](https://support.anthropic.com/)
- [Console per sviluppatori](https://console.anthropic.com/)
- [Discord per Sviluppatori](https://www.anthropic.com/discord)
- [Supporto](https://support.anthropic.com/)
- [開発者コンソール](https://console.anthropic.com/)
- [開発者向けDiscord](https://www.anthropic.com/discord)
- [サポート](https://support.anthropic.com/)
- [Consola de Desarrollador](https://console.anthropic.com/)
- [Discord para Desarrolladores](https://www.anthropic.com/discord)
- [Soporte](https://support.anthropic.com/)
- [개발자 콘솔](https://console.anthropic.com/)
- [개발자 Discord](https://www.anthropic.com/discord)
- [지원](https://support.anthropic.com/)
- [开发者控制台](https://console.anthropic.com/)
- [开发者 Discord](https://www.anthropic.com/discord)
- [支持](https://support.anthropic.com/)
- [開發者控制台](https://console.anthropic.com/)
- [開發者 Discord](https://www.anthropic.com/discord)
- [支援](https://support.anthropic.com/)
- [Консоль разработчика](https://console.anthropic.com/)
- [Дискорд для разработчиков](https://www.anthropic.com/discord)
- [Поддержка](https://support.anthropic.com/)
- [Konsol Pengembang](https://console.anthropic.com/)
- [Discord Pengembang](https://www.anthropic.com/discord)
- [Dukungan](https://support.anthropic.com/)
- [Console do Desenvolvedor](https://console.anthropic.com/)
- [Discord para Desenvolvedores](https://www.anthropic.com/discord)
- [Suporte](https://support.anthropic.com/)