| Benchmark | Claude Fable 5 | Kimi K2 (Reasoning) |
|---|---|---|
| AA Agentic Index | 56.6 | 47.9 |
| AA Intelligence | 62.1 | 33.5 |
| AA-LCR | 76.7 | 70.3 |
| AA-Omniscience | 43.3 | -21.4 |
| Artificial Analysis Coding Index | 76.5 | 34.8 |
| browsecomp | 88 | 60.2 |
| critpt | 28.6 | 2.6 |
| gdpval | 61.9 | 24.5 |
| GPQA Diamond | 92.6 | 84.5 |
| HLE | 64.5 | 23.9 |
| HLE (with tools) | 64.5 | 44.9 |
| IFBench | 63.5 | 68.1 |
| OmniScience Accuracy | 65.3 | 30.9 |
| OmniScience Non-Hallucination | 36.4 | 25.8 |
| scicode | 60.2 | 44.8 |
| simplebench | 81.9 | 39.6 |
| SWE-bench Multilingual | 86.6 | 61.1 |
| SWE-bench Verified | 95 | 71.3 |
| Terminal-Bench Hard | 62.9 | 31.1 |
| τ²-Bench Telecom (AA run) | 98.5 | 93 |
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.