Claude Fable 5.1
Anthropic · Flagship · Sep 1, 2026New
Anthropic's most capable generally available model, built for the hardest reasoning and long-running agent work. Thinking is always on.
- Artificial Analysis Intelligence Index
- 53.4#3 of 29 · max effort
- API price per 1M tokens
- $10 / $50input / output · $20 blended
- Context window
- 1Mup to 128K output tokens
- Output speed
- 66 t/s4m 54s to first answer at max effort
Benchmark results
Each row shows this model (blue) against every other model in the catalogue (grey), worst on the left, best on the right.
Overall capability
- Artificial Analysis Intelligence Index53.4#3 of 29
Max effort, default refusal fallback enabled
- Epoch Capabilities Index164.7#5 of 16
- LiveBench83.4%#1 of 23
Max effort
Coding
- Terminal-Bench 4.057.9%#5 of 13
Claude Code harness, max effort
- SWE-Bench Pro (V2 Hard)92.2%#1 of 5
Claude Code harness, high effort
- Arena WebDev1,745#5 of 27
Max effort
Agents & professional work
- GDPval-AA1,756#3 of 18
- Vending-Bench 2$5,422#6 of 9
Reasoning
- ARC-AGI-290%#4 of 11
Max effort
- FrontierMath (Tiers 1–3)90.2%#4 of 14
Max effort
- MathArena82.6%#4 of 10
Max effort
- SimpleBench86.6%#2 of 12
Knowledge & science
- Humanity's Last Exam (Diamond)51.3%#3 of 9
Multimodal
- Arena Vision1,288#5 of 15
Max effort
Human preference
- Arena Text1,501#3 of 24
Max effort
Reasoning effort trade-offs
The same model at different thinking budgets. More effort buys a higher intelligence score at a higher cost per task and a longer wait. Measured by Artificial Analysis.
| Effort | Score | Cost / task | First answer | Speed |
|---|---|---|---|---|
| Low | 46.8 | $2.37 | 3.9s | 54 t/s |
| Medium | 48.9 | $2.98 | 7.6s | 54 t/s |
| High | 51.1 | $3.91 | 23s | 58 t/s |
| Xhigh | 53.2 | $5.98 | 81s | 58 t/s |
| Max (headline score) | 53.4 | $7.63 | 4m 54s | 66 t/s |
Specs & pricing
- Developer
- Anthropic
- Released
- Sep 1, 2026
- Availability
- Generally available
- API model ID
claude-fable-5-1- Weights
- Proprietary (API only)
- Reasoning
- Always on
- Knowledge cutoff
- Jun 2026
- Accepts
- Text, Image
- Produces
- Text
- Context window
- 1,000,000 tokens
- Max output
- 128,000 tokens
- Input price
- $10 / 1M tokens
- Output price
- $50 / 1M tokens
- Cached input
- $0.25 / 1M tokens
Closest alternatives
Models from other labs nearest to Claude Fable 5.1 on Artificial Analysis Intelligence Index.
GPT-6 Astra
OpenAI
Intelligence 52.7 · $10 / $50
Compare side by sideGemini 4 Argon
Google
Intelligence 52.6 · $2 / $10
Compare side by sideGPT-6.1 Sol
OpenAI
Intelligence 51.8 · $2 / $10
Compare side by side