Claude Fable 5.1
Claude Fable 5.1
Anthropic's top tier for the longest, hardest runs, with a cheaper Opus now snapping at its heels.
Updated
The read
Fable 5.1 is Anthropic's most expensive model and the one it positions for demanding reasoning and long-horizon agent work. It is a clear step up from Fable 5 and cuts the price of cached input by three quarters, which matters on long runs that re-read the same context. The trouble is Opus 5.5, which arrived three weeks later, scores higher on both independent leaderboards and costs well under half as much. Fable now earns its place only on work where you have measured the difference.
Where it shines
- Built and priced for hours-long agent runs that must not drift
- Cached input costs a quarter of what it did on Fable 5
- Strong on computer-use and terminal benchmarks
Worth knowing first
- Outscored by Opus 5.5 at well under half the price
- Slow and expensive enough to be the wrong default for interactive work
Index score
87.9
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 08 of 73
- Builds
- LLMs
- Category
- Frontier
- Output
- Text, code and tool calls over an API
- Pricing
- Paid only
- Best for
- Long autonomous runs where you have measured that the top tier pays off
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 01Claude Opus 5.5Claude Opus 5.5ProvisionalFable-level answers for less than the Opus it replaces. The new default.Frontier modelPaid only92.8
- 03Gemini 4 ArgonGemini 4 ArgonProvisionalTop of the human-preference leaderboard, and almost nobody can use it yet.Frontier modelPaid only90.6
- 04GPT-6.1 SolGPT-6.1 SolProvisionalNear-Astra answers at a fifth of the price. The OpenAI model most teams should deploy now.Frontier modelPaid only90.3
