Claude Fable 5 vs Claude Sonnet 5.5
Vals Science Index?
Fable 5—
Sonnet 5.559.1
| Benchmark | Fable 5 | Sonnet 5.5 |
|---|---|---|
| Humanity's Last Exam | 55.5 | 45.8 |
| Arena Text · overall | 1,504 · high | 1,475 · xhigh |
| Arena Text · Chinese | 1,542.3 · high | 1,536.9 · xhigh |
| Arena Text · English | 1,499 · high | 1,484.5 · xhigh |
| Arena Text · French | 1,503.6 · high | 1,496.9 · xhigh |
| Arena Text · instruction following | 1,502.3 · high | 1,488.5 · xhigh |
| Arena Text · non-English | 1,480.7 · high | 1,454.2 · xhigh |
| Arena Text · Russian | 1,503 · high | 1,445.2 · xhigh |
| Arena Text · Spanish | 1,495.8 · high | 1,450.2 · xhigh |
| AA Omniscience Index | 43.3 | 21 |
| SimpleQA Verified v1 | 70.7 · xhigh | 46.5 · max |
| CritPt (physics) | 28.6 | 24.6 |
| AA Intelligence Index 4.3.2 | 49.6 | 46.8 |
| AA Omniscience · Accuracy | 65.4 | 52 |
| AA Omniscience · Hallucination Rate | 63.6 | 64.6 |
| AA-LCR (long context) 1.1 | 82.3 | 78 |
| GDP.pdf · AA all-pass | 24 | 25.2 |
| MLCR-AA | 64.4 | 75 · max |
| Chess Puzzles | 47 | 59 · max |
| FrontierMath Erdos | 0 | 2.9 · max |
| FrontierMath Tier 4 (v2) | 90.2 | 80.5 · max |
| FrontierMath Tiers 1–3 (v2) | 87 | 88.8 · max |
| Mystery Game Puzzles | 52 | 65 · max |
| Vals MedCode | 56.1 | 52.9 |
| Vals MedScribe | 88.5 | 91.1 |
| Vals Legal Research Bench | 49.5 | 48.1 |
| Arena Text · German | 1,486.7 · high | — |
| Arena Text · Japanese | 1,506.4 · high | — |
| Arena Text · Korean | 1,491 · high | — |
| Arena Text · Polish | 1,493.4 · high | — |
| IFBench (instructions) | 63.5 | — |
| ARC-AGI-1 · public evaluation | 97.6 | — |
| ARC-AGI-1 · semi-private | 98.5 | — |
| ARC-AGI-2 · public evaluation | 96.4 | — |
| ARC-AGI-2 · semi-private | 89.2 | — |
| EQ-Bench 3 | 2,049.7 | — |
| MathArena · ArXivMath · April 2026 | 72.5 | — |
| MathArena · ArXivMath · June 2026 | 85.4 | — |
| MathArena · ArXivMath · May 2026 | 87.5 | — |
| MathArena · BrokenArXiv · April 2026 | 54.1 | — |
| MathArena · BrokenArXiv · June 2026 | 47.8 | — |
| MathArena · BrokenArXiv · May 2026 | 44.5 | — |
| MathArena · ArXivLean · June 2026 | 15.2 | — |
| Vals LegalBench | 88.6 | — |
| Vals Science Index | — | 59.1 |
Specs
| Spec | Fable 5 | Sonnet 5.5 |
|---|---|---|
| Input, per 1M | $10 | $2 |
| Output, per 1M | $50 | $10 |
| Cost per task? | $8.75 | $0.88 |
| Speed, tok/s | 67 | 106 |
| Context | 1M | 1M |
| Weights | Closed | Closed |
| Released | Jun 2026 | Sep 2026 |
| See images | ✓ | ✓ |
| Read PDFs | Partial | ✓ |
| Hear audio | — | — |
| Watch video | — | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | — |
| Return strict JSON | ✓ | ✓ |
| Search the web | ? | ✓ |
| Run code | ? | ✓ |
| Use a computer | ? | ✓ |
What breaks when you switch
Fable 5 → Sonnet 5.5
No forced tool use›
Forced tool_choice any or tool returns HTTP 400; remove the forced choice.
Replace disabled thinking›
Disabled thinking returns HTTP 400; use between_tools for the lowest thinking setting.
No manual thinking budgets›
Manual thinking budgets and extra between_tools fields return HTTP 400; remove manual budgets and send between_tools without display, budget_tokens or block_binding.
Remove sampling parameters›
Non-default temperature, top_p or top_k returns HTTP 400; remove these parameters.
No assistant prefill›
A prefilled final assistant turn returns HTTP 400; remove that prefill.
Old computer tool rejected · Claude API / Google Cloud›
The computer_20251124 tool returns HTTP 400 on the Claude API and Google Cloud; use computer_toolset_20260801 there.
Sonnet 5.5 → Fable 5
Nothing known.