| Benchmark | Kimi K2.6 | Muse Spark |
|---|---|---|
| AA Agentic Index | 58.7 | 28.7 |
| AA Intelligence | 45.1 | 44.3 |
| AA-LCR | 76.7 | 77 |
| AA-Omniscience | 6.4 | 7.2 |
| arena_elo | 1461 | 1488 |
| arena_text_factuality | 1455 | 1467 |
| arena_vision | 1264 | 1295 |
| Artificial Analysis Coding Index | 61.8 | 58.6 |
| coding_arena_elo | 1509 | 1540 |
| critpt | 8 | 11.3 |
| deepsearchqa_f1 | 92.5 | 74.8 |
| frontiermath_tier_4 | 14.6 | 14.6 |
| gdpval | 41.4 | 32.3 |
| GPQA Diamond | 91.1 | 89.5 |
| HLE | 37.5 | 58.4 |
| IFBench | 76 | 75.9 |
| itbenchSre | 31.2 | 16.2 |
| MCP Atlas | 68.1 | 82.2 |
| MMMU-Pro | 80.1 | 80.5 |
| OmniScience Accuracy | 32.8 | 49.6 |
| OmniScience Non-Hallucination | 60.7 | 26.8 |
| scicode | 53.5 | 51.5 |
| simpleqa_verified | 38.7 | 66.3 |
| SWE-bench Pro | 58.6 | 55 |
| SWE-bench Verified | 80.2 | 77.4 |
| TauBench V3 - Banking | 23.3 | 19.6 |
| Terminal-Bench 2.0 | 66.7 | 59 |
| Terminal-Bench 2.1 | 65.9 | 62.2 |
| Terminal-Bench Hard | 43.9 | 45.5 |
| τ²-Bench Telecom (AA run) | 95.9 | 91.5 |
| τ³-Bench | 20.6 | 19.6 |
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.