Claude Fable 5.1 Text Generation API

anthropic/claude-fable-5.1
1,000,000 tokens · 1600 input credits / 1M tokens

Claude Fable 5.1 Text Generation turns text and image context into code, research findings, and detailed answers for long-running work. Its 1M-token context window and adaptive reasoning help carry requirements and evidence across connected steps.

Claude Fable 5.1

Anthropic · chat-completions

Chat with Claude Fable 5.1

Each model has its own conversation. Switching never sends another model's history, and switching back resumes where you left off. Requests are billed from actual token usage.

Ctrl / ⌘ + Enter to send0 / 32,000

Continue with

Claude Fable 5.1 Text Generation

Claude Fable 5.1 works across lengthy coding projects, multistep research, and document-heavy analysis. Send ordered conversation messages to Claude Messages or Chat Completions and receive generated content and usage as JSON or an SSE stream.

Why Choose Claude Fable 5.1?

  • Long-running codingTrack requirements, source files, and verification results through multi-file features, refactors, debugging, and code review.

  • Multistep researchCarry evidence through connected research questions and turn the findings into a structured report or other finished work.

  • Document and image understandingRead dense charts, tables, and diagrams alongside text to extract details that matter to an analysis.

  • Scientific and mathematical reasoningWork through multi-step mathematics and technical questions while keeping the assumptions and conclusion tied to the problem.

  • Long-form writingProduce readable, well-organized prose that follows a specified style across extended writing tasks.

Parameters

ParameterRequirementDescription
modelRequired

Use anthropic/claude-fable-5.1 for both protocols.

messagesRequired

A non-empty ordered array of role and content messages.

max_tokensRequired for Claude Messages

A positive integer that sets the output-token limit. Chat Completions accepts this field optionally.

systemOptional for Claude Messages

System instructions as a string or structured text blocks.

cache_controlOptional for Claude Messages

Places a prompt-cache marker on cacheable request content.

streamOptional

Use false for a JSON response or true for an SSE event stream.

Defaultfalsetrue

How to Use

  1. State the taskDescribe the target result, constraints, source material, and review criteria.

  2. Choose the protocolUse Claude Messages for Anthropic-compatible requests or Chat Completions for OpenAI-style clients.

  3. Set the output limitInclude a positive max_tokens value for Claude Messages and send the ordered messages array.

  4. Review the responseRead the generated content and usage, then supply new evidence or instructions for the next turn.

Pricing

Vidgo AI lists separate rates for input, prompt-cache reads, and output. Each rate below applies per 1M tokens.

UsageRateDetails
Input$8.00 · 1600 credits / 1M tokensStandard input tokens in the request.
Cache read$0.20 · 40 credits / 1M tokensInput tokens read from prompt cache.
Output$40.00 · 8000 credits / 1M tokensTokens generated in the response.

Best Use Cases

  • Codebase-wide featuresUse repository files, architecture constraints, and tests to plan and review connected changes.

  • Research synthesisCombine reports, source excerpts, and competing evidence into a documented finding.

  • Document analysisExamine text alongside charts, tables, and diagrams to extract details and relationships.

Pro Tips

  • Provide the intended deliverable, source evidence, and acceptance criteria together for long-running tasks.
  • Keep recurring instructions and reference material stable when using prompt caching.
  • For document analysis, pair image blocks with a specific question about the chart, table, or diagram.

Usage Notes

  • The model has a 1,000,000-token context window and a 128,000-token maximum output.
  • The on-page Playground sends Chat Completions requests and offers an output control from 256 to 8,192 tokens.
  • Use anthropic/claude-fable-5.1 with /v1/messages or /v1/chat/completions.

Related Models

Claude Fable 5.1 Text Generation API frequently asked questions

What is the Claude Fable 5.1 Text Generation API?

Claude Fable 5.1 Text Generation is an Anthropic model for long-running coding, research, and complex knowledge work. It turns text and image context into code, analysis, and detailed answers while tracking requirements across connected steps. Its 1M-token context window and adaptive reasoning support extensive source material and multistep work. You can call it programmatically or try it from the playground above.

How does Claude Fable 5.1 handle long-running coding?

Claude Fable 5.1 can work across connected coding tasks, including multi-file features, refactors, debugging, and review. Provide the relevant repository files, target behavior, and test results so each step can be checked against the same requirements.

How much context can Claude Fable 5.1 analyze?

Claude Fable 5.1 has a 1,000,000-token context window. Include the relevant documents, source files, and instructions in the conversation when the task depends on their relationships.

How many output tokens can Claude Fable 5.1 generate?

The model specification allows up to 128,000 output tokens. Set max_tokens for the API response; the on-page Playground provides a 256–8,192-token control for interactive trials.

How does Claude Fable 5.1 interpret charts and diagrams?

Claude Fable 5.1 combines image and text input to analyze charts, diagrams, tables, and dense documents. In Claude Messages, place supported image blocks in messages.content and ask a specific question about the visual evidence.

How does Claude Fable 5.1 use adaptive reasoning?

Claude Fable 5.1 uses adaptive thinking with high effort by default. Describe the decision, constraints, and evidence clearly so the model can devote its reasoning to the parts of the task that need it.

How can Claude Fable 5.1 reuse prompt context?

Claude Fable 5.1 can read cached prompt content across repeated requests. Use cache_control on stable instructions or reference material in Claude Messages, then keep that content unchanged when continuing the workflow.