Claude Fable 5 · $5 / $25 per 1M tokens · 1M context
Claude Fable 5 API
Claude-compatible access to anthropic/claude-fable-5 with half-price tokens, 1M context, 128K max output, streaming, and drop-in Messages API support.
Half the official API price
$5 input / $25 output per 1M tokens
Official Claude Fable 5 pricing is $10 input and $50 output per 1M tokens. Claude Fable 5 API cuts both sides in half while keeping Claude-compatible request semantics.
What you can build
Use Claude Fable 5 API for products and internal tools that need more than short chat: large context, long outputs, vision, and durable multi-step reasoning.
Coding assistants and repo analysis
Load architecture notes, source files, tests, logs, and issue context together. Ask for bug diagnosis, refactors, migration plans, test generation, documentation updates, or code review across an entire repository.
Long-document research
Analyze contracts, filings, technical papers, product specs, meeting transcripts, or multi-file knowledge bases. The 1M-token context helps preserve evidence instead of forcing early summarization.
Visual reasoning workflows
Send screenshots, charts, diagrams, UI mockups, and image-heavy reports. Combine visual evidence with text instructions to debug interfaces, interpret charts, or reason over spatial layouts.
Long-running agents
Build agents that plan, call tools, read results, update state, and continue. Claude Fable 5 is useful for workflows such as research automation, software maintenance, data cleanup, and operations assistance.
What is Claude Fable 5 API?
A lower-cost Claude-compatible API for Anthropic's latest generally available Mythos-class model
Claude Fable 5 API gives developers access to Claude Fable 5 through a Claude-compatible Messages API. The model is positioned for demanding professional workloads: large software projects, multi-file refactors, repository analysis, long-form research, image and chart reasoning, and autonomous agents that need to plan, execute, inspect results, and continue across many steps. You get a 1,000,000-token context window, support for text and image inputs, up to 128,000 output tokens for long responses, standard JSON responses, and live streaming with server-sent events. Pricing is $5 input and $25 output per 1M tokens through this API, half of the official $10 / $50 rate, while preserving the Claude request shape your tools already use.
Half the official price: $5 input and $25 output per 1M tokens, compared with the official $10 / $50 rate
1M-token context: load whole repositories, long documents, chat histories, images, and multi-step agent state without aggressive trimming
Up to 128K output tokens: generate long reports, migration plans, technical specs, and large code changes in fewer calls
Drop-in compatible: keep Claude-style model, messages, max_tokens, temperature, top_p, stop, stream, and user fields
Built for long-horizon agents: useful for workflows that need planning, execution, tool feedback, self-checking, and continuation
Streaming built in: choose one complete JSON response or live server-sent events for interactive products
Live in four steps
Switch to Claude Fable 5 API without rewriting your Claude code
Step 1
Grab your DefAPI key and send it as Authorization: Bearer <your-api-key>
Step 2
Point your base URL to https://api.defapi.org/api/v1/messages
Step 3
Set model to anthropic/claude-fable-5 and add your messages array
Why Claude Fable 5 API
More than a cheap endpoint: Claude Fable 5 combines long context, strong coding ability, visual reasoning, and agent-ready behavior behind a familiar Claude-style API.
Half the Official Price
$5 input and $25 output per 1M tokens, half the official $10 / $50. Keep the same Claude Fable 5 capability while reducing the cost of high-volume coding, research, and automation workloads.
Drop-In Claude Compatibility
Use the Claude Messages API format your application already understands. Point your base URL to the DefAPI endpoint, keep your message array and generation parameters, and migrate with a minimal code change.
1M-Token Context Window
Give the model complete project context: source files, architecture notes, logs, long PDFs, chat history, screenshots, and task state. This is especially useful for repo-wide refactors and document-heavy analysis.
Agent-Ready Autonomy
Claude Fable 5 is designed for long-horizon work. It can break down complex goals, continue through multi-stage tasks, inspect intermediate results, and support assistants that operate for hours instead of one prompt at a time.
Up to 128K Output Tokens
Long outputs matter when you need complete implementation plans, detailed audits, multi-section research reports, structured JSON, or large code patches. Fewer forced continuations means simpler application logic.
Vision and Spatial Reasoning
Use image inputs for screenshots, charts, diagrams, UI states, and visual debugging. Claude Fable 5 is especially useful when a task mixes text, code, and visual context.
Adaptive Thinking for Hard Tasks
The model is built to spend more effort on harder requests, making it a strong fit for ambiguous debugging, complex reasoning, multi-document synthesis, and tasks where shallow answers are not enough.
Clear Safety Boundaries
Claude Fable 5 is the broadly available, safety-classified version of the Mythos-class model family. Some high-risk requests may be refused or routed through fallback behavior, which helps keep powerful capability usable in public products.
Frequently Asked Questions
API compatibility, pricing, model capability, context length, output limits, safety behavior, and practical use cases for Claude Fable 5 API.
What is Claude Fable 5 API?
Claude Fable 5 API gives you Claude Fable 5 through a Claude-compatible Messages endpoint at https://api.defapi.org/api/v1/messages. It is designed as a drop-in API option for developers who want frontier Claude capability, a 1M-token context window, streaming support, and lower per-token pricing.
What is Claude Fable 5?
Claude Fable 5 is Anthropic's latest generally available Mythos-class model. It is positioned for demanding tasks such as software engineering, long-horizon reasoning, knowledge work, visual understanding, and autonomous agent workflows.
How is Claude Fable 5 related to Claude Mythos 5?
According to the project notes, Fable 5 uses the same underlying model family as Mythos 5 but adds stricter safety classifiers for broad public availability. In some high-risk areas, it may refuse a request or use fallback behavior.
Which model ID do I use?
Use anthropic/claude-fable-5 in the model field of your request body when calling the DefAPI Messages endpoint.
How do I authenticate?
Send your API key in the Authorization header as a Bearer token, for example: Authorization: Bearer <your-api-key>. Keep the key on your server for production apps and avoid exposing it in client-side code.
How much does Claude Fable 5 API cost?
This API prices Claude Fable 5 at $5 input and $25 output per 1M tokens, half of the official $10 / $50 rate listed in the source notes. This is especially valuable for long-context and agent workloads where token volume can grow quickly.
What is the context window?
Claude Fable 5 supports a 1,000,000-token context window. That makes it practical to include large codebases, long documents, images, task history, logs, and intermediate agent state in a single request.
What is the maximum output length?
The source notes list up to 128,000 output tokens. This is useful for long technical reports, detailed code migration plans, structured outputs, and large generation tasks that would otherwise require many continuation calls.
Does Claude Fable 5 API support streaming?
Yes. Set stream to true for server-sent event streaming, or stream to false for one complete JSON response. Streaming is useful for chat UIs, coding assistants, and agent consoles where users should see progress as it happens.
Is it compatible with Claude-style requests?
Yes. The endpoint accepts Claude-style fields such as model, messages, max_tokens, temperature, top_p, stop, stream, and user. That makes it suitable for applications already built around the Claude Messages API shape.
Can I send images or visual context?
Claude Fable 5 is described as strong in vision and spatial reasoning. Use it for screenshots, UI states, charts, diagrams, visual bug reports, and tasks that require combining image evidence with text instructions.
What can I build with Claude Fable 5 API?
Common use cases include coding assistants, repository analysis, document reasoning, research agents, chart and image understanding, automated code review, long-running workflow automation, and internal tools that need large context.
What limitations should I expect?
Claude Fable 5 can be slower and more expensive than smaller models for simple tasks, especially when prompts are large or reasoning effort is high. Its safety classifiers may also refuse or redirect some high-risk requests, including certain cybersecurity, biological, or chemical topics.
What errors might the API return?
Common errors include 400 for invalid parameters, 401 for a missing or invalid API key, and 500 for a server error. If a request is blocked by model safety behavior, handle the refusal path separately from transport or authentication errors.
Is this an official Anthropic website?
No. This is an independent community resource for Claude Fable 5 API information and API access. It is not official and not affiliated with Anthropic.