← All models

Claude Fable 5

Published Publisher Anthropicclaude-fable-5

Claude Fable 5

Positioning and release

Claude Fable 5 is Anthropic’s general-use, Mythos-class model for complex coding and knowledge work. It was released on June 9, 2026. Access was suspended on June 12 and restored on July 1, 2026. Anthropic currently lists it as active (legacy), with retirement no sooner than June 9, 2027. (anthropic.com)

Capabilities and uses

  • Handles long-running, agentic software engineering and analytical tasks, including codebase changes, document analysis, and interpretation of charts and tables.
  • Accepts text and images and produces text. Anthropic describes vision uses such as extracting information from scientific figures and reconstructing web interfaces from screenshots.
  • Has a 1 million-token context window, a 128,000-token maximum output, and always-on adaptive thinking. Its documented knowledge and training-data cutoffs are January 2026. (anthropic.com)

Limitations and safeguards

Outputs can be incorrect and should be reviewed, especially for consequential decisions or autonomous work. Anthropic says its safety classifiers can mistakenly flag benign requests. At launch, certain cybersecurity, biology, chemistry, and model-distillation requests could be routed to Claude Opus 4.8 instead; safeguard behavior may change as it is updated. Claude Mythos 5 is a separately offered, restricted model and should not be confused with Fable 5. (anthropic.com)

Claude API pricing

Anthropic’s list prices below are in USD per million tokens, apply to all context-window sizes, and are specific to Claude Fable 5 on the Claude API. Cache write and read prices apply to the corresponding cached input tokens; output is billed separately. Global and US-only inference are distinct scopes. Prices on other platforms may differ. The rate card is effective November 9, 2026; checked November 16, 2026. (assets.claude.com)

Tier and scope Input Output 5-minute cache write 1-hour cache write Cache read
Standard, global $10 $50 $12.50 $20 $1
Standard, US-only inference $11 $55 $13.75 $22 $1.10
Batch, global $5 $25 $6.25 $10 $0.50
Batch, US-only inference $5.50 $27.50 $6.875 $11 $0.55

Official sources