Claude Fable 5.2
claude-fable-5-2
Anthropic · reasoning · api
Context
1M
Max output
128K
Input $/1M
$8.00
Output $/1M
$40.00
Modalities
text · image · document
Released
09 Nov 2026
AI summary
● machine-written
Claude Fable 5.2 — reasoning model for demanding agentic work
Claude Fable 5.2 is Anthropic's most capable reasoning model, designed for demanding long-horizon agentic reasoning and coding tasks. It processes up to 1M input tokens and supports text, image, and document inputs with adaptive thinking enabled by default. Pricing is $8 per million input tokens and $40 per million output tokens.
What's new
- Adaptive thinking enabled by default for reasoning tasks
- Supports text, image, and document inputs
- 1M token context window with 128K output token limit
- Available across Claude API, Google Cloud (Anthropic-managed and Google-managed), AWS, and Microsoft Foundry
Best for
Long-horizon agentic work requiring complex reasoningDemanding reasoning tasks where other models fall shortAutonomous knowledge work and planningCoding tasks requiring extended context
Common questions about Claude Fable →
Source: https://docs.claude.com/en/docs/about-claude/models/overview