Scan to join WeChat groupClaude Fable 5.1
Claude Fable 5.1 for complex coding and long-running agents, with 75% lower cache-read pricing
Released on September 1, 2026, Claude Fable 5.1 is Anthropic’s generally available model for complex coding, knowledge work and long-running agents. Its API model ID is claude-fable-5-1. Standard input and output pricing remains $10 and $50 per million tokens, while cache reads cost $0.25 per million tokens.
starsCapabilities
paymentsContext and pricing
descriptionOverview
Overview
Claude Fable 5.1 was released on September 1, 2026, with API model ID claude-fable-5-1. It targets cross-codebase development, complex knowledge work and long-running agent tasks.
Anthropic highlights improvements in root-cause analysis, multi-step execution, self-verification and progress updates. Vision supports understanding charts, documents and PDFs, as well as checking generated work against design goals.
Availability and pricing
Fable 5.1 is available to Pro, Max, Team and Enterprise users, and through the Claude Platform, AWS, Google Cloud and Microsoft Foundry. Integration details depend on the platform.
As verified on September 14, 2026, standard API pricing is $10 per million input tokens, $50 per million output tokens and $0.25 per million cache-read tokens. Cache reads cost 75% less than Fable 5; this is not a 75% reduction in every request’s total cost. Actual costs depend on cache hits and input/output usage.
Use cases and model boundaries
Consider it for code development and review, long-running agent workflows, document analysis and multi-stage enterprise work. These are the use cases described by Anthropic; evaluate performance on your own tasks.
Fable 5.1 and Mythos 5.1 share the same underlying model but have different safeguards and access conditions. Fable 5.1 is generally available, while Mythos 5.1 is offered through trusted access programs. They are not interchangeable API aliases.
lightbulbUse cases
- Cross-codebase development, code review and root-cause analysis
- Long-running agents and multi-tool workflows
- Charts, PDFs and complex document analysis
- Multi-stage enterprise knowledge work
thumb_upStrengths
- Designed for complex coding and long-running tasks
- Emphasis on self-verification and progress updates
- Vision-assisted document understanding and output checks
- 75% lower cache-read pricing than Fable 5
infoLimitations
- Actual costs depend on cache hits and output usage
- Long-running reliability needs workload-specific evaluation
- Access conditions and safeguards differ from Mythos 5.1