Claude Fable 5.1 vs Claude Opus 5
Vals Science Index?
Fable 5.1—
Opus 547.9
| Benchmark | Fable 5.1 | Opus 5 |
|---|---|---|
| Humanity's Last Exam | 53.8 | 51.3 |
| Arena Text · overall | 1,502 · max | 1,490 · max |
| Arena Text · Chinese | 1,585.6 · max | 1,573.4 · max |
| Arena Text · English | 1,504.9 · max | 1,507.3 · max |
| Arena Text · French | 1,527.3 · max | 1,519.5 · high |
| Arena Text · German | 1,501.7 · max | 1,525.9 · max |
| Arena Text · instruction following | 1,517.2 · max | 1,517.8 · max |
| Arena Text · Japanese | 1,560.1 · max | 1,515.4 · max |
| Arena Text · Korean | 1,520.8 · max | 1,520.7 · max |
| Arena Text · non-English | 1,507.5 · max | 1,500.5 · max |
| Arena Text · Polish | 1,524.3 · max | 1,519.9 · max |
| Arena Text · Russian | 1,518.6 · max | 1,507.2 · max |
| Arena Text · Spanish | 1,520.5 · max | 1,519.5 · max |
| AA Omniscience Index | 37.6 | 31 |
| SimpleQA Verified v1 | 70.8 · max | 63.3 |
| CritPt (physics) | 29.1 | 26.9 |
| AA Intelligence Index 4.3.2 | 48.9 | 44.8 |
| AA Omniscience · Accuracy | 63.1 | 57.1 |
| AA Omniscience · Hallucination Rate | 69.1 | 60.7 |
| AA-LCR (long context) 1.1 | 84.7 | 82 |
| GDP.pdf · AA all-pass | 26.8 | 20 |
| MLCR-AA | 71.1 · max | 56.1 |
| ARC-AGI-1 · public evaluation | 96.6 | 99 · max |
| ARC-AGI-1 · semi-private | 94.5 | 97.5 · max |
| ARC-AGI-2 · public evaluation | 87.5 | 97.1 · max |
| ARC-AGI-2 · semi-private | 86.3 | 90.4 · max |
| Chess Puzzles | 48 · max | 43 · max |
| FrontierMath Erdos | 0 · max | 0 · max |
| FrontierMath Tier 4 (v2) | 87.8 · max | 73.2 · max |
| FrontierMath Tiers 1–3 (v2) | 90.2 · max | 85.6 · max |
| Mystery Game Puzzles | 58 · max | 59 · max |
| MathArena · ArXivMath · June 2026 | 91 · max | 82.6 · max |
| MathArena · BrokenArXiv · June 2026 | 86.6 · low | 90.7 · max |
| Vals LegalBench | 88.5 | 87 |
| Vals MedCode | 53.5 | 63.6 |
| Vals MedScribe | 91.3 | 91 |
| Vals Legal Research Bench | 55.3 | 55.3 |
| ARC-AGI-3 · semi-private | — | 30.2 · high |
| MathArena · ArXivMath · April 2026 | 77.5 · max | — |
| MathArena · ArXivMath · August 2026 | 87.7 · high | — |
| MathArena · ArXivMath · May 2026 | 87.5 · max | — |
| MathArena · BrokenArXiv · August 2026 | 79.8 · high | — |
| MathArena · BrokenArXiv · May 2026 | 78 · low | — |
| MathArena · ArXivLean · June 2026 | — | 32.6 · max |
| Vals Science Index | — | 47.9 |
Specs
| Spec | Fable 5.1 | Opus 5 |
|---|---|---|
| Input, per 1M | $10 | $5 |
| Output, per 1M | $50 | $25 |
| Cost per task? | $2.98 | $2.19 |
| Speed, tok/s | 56 | 55 |
| Context | 1M | 1M |
| Weights | Closed | Closed |
| Released | Sep 2026 | Jul 2026 |
| See images | ✓ | ✓ |
| Read PDFs | Partial | Partial |
| Hear audio | — | — |
| Watch video | — | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | Partial |
| Return strict JSON | ✓ | ✓ |
| Search the web | ? | ? |
| Run code | ? | ? |
| Use a computer | ? | ? |
What breaks when you switch
Fable 5.1 → Opus 5
Nothing known.
Opus 5 → Fable 5.1
Replace forced tool_choice›
Forced tool_choice returns HTTP 400; use auto with an explicit instruction and strict tools, or JSON outputs.
Adaptive thinking is always on›
Disabled thinking or manual budget_tokens returns HTTP 400; omit thinking or use adaptive thinking.
No non-default sampling values›
Non-default temperature, top_p or top_k returns HTTP 400; omit non-default sampling values.
Requires 30-day data retention · Claude API›
A Claude API organization or workspace without 30-day retention receives HTTP 400; configure workspace retention or contact the Anthropic account team.
Thinking blocks bind to conversations›
Where conversation binding is enforced, replaying an invalidated block returns HTTP 400; use the documented drop_block control to continue without it.
No assistant prefill›
Assistant prefill returns HTTP 400; use system prompt instructions instead.