Claude Fable 5.1 scores 69.4 on the Respan Index (#4), ahead of DeepSeek V4 Pro at 46.9 (#29). Claude Fable 5.1 leads on 20 of the 21 benchmarks both report. DeepSeek V4 Pro is 10x cheaper per token at list price.
* Estimated: no published scores in that category. How the Respan Index works
| LiveBench Agentic Coding (LiveBench) |
| 66.1% |
| 54.9% |
| LiveBench Coding (LiveBench) | 86.4% | 77.2% |
| LiveBench Data Analysis (LiveBench) | 80.3% | 79.2% |
| LiveBench Instruction Following (LiveBench) | 73% | 67.7% |
| LiveBench Language (LiveBench) | 89.5% | 82.1% |
| LiveBench Mathematics (LiveBench) | 97% | 95.1% |
| LiveBench Reasoning (LiveBench) | 91.7% | 85.8% |
| Mystery Game Puzzles (Epoch AI) | 58% | 43% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 100% | 96.7% |
| SimpleQA Verified | 70.8% | 47% |
| Terminal-Bench 4.0 (Vals AI) | 58.1% | 14.1% |
| Vending-Bench 2 (Andon Labs) | 5421.56 | 3284.52 |
| WeirdML (Håvard Tveit Ihle) | 92.9% | 66.2% |
| LMArena Agent (LMArena) | 0.1431 | 0.0119 |
| LMArena Elo | 1501 | 1451 |
| LMArena WebDev (LMArena) | 1749 | 1583 |
| SimpleBench (SimpleBench) | 86.6% | 50.9% |
Higher is better. Each score links to where it was published.