| Benchmark | Kimi K3 | Qwen3.7 Max |
|---|---|---|
| AA Agentic Index | 54.3 | 30.9 |
| AA Intelligence | 60 | 46.7 |
| AA-LCR | 82.7 | 74.7 |
| AA-Omniscience | 46 | 13.5 |
| aime_2026 | 96.7 | 97 |
| arena_elo | 1489 | 1474 |
| arena_text_factuality | 1471 | 1479 |
| Artificial Analysis Coding Index | 76.2 | 66 |
| coding_arena_elo | 1674 | 1517 |
| critpt | 23.4 | 13.4 |
| DeepSWE | 67.5 | 18 |
| DeepSWE 1.1 | 69 | 21.6 |
| Finance Agent v2 | 54.4 | 48.4 |
| gdpval | 59.4 | 38.5 |
| GPQA Diamond | 93.5 | 92.4 |
| HLE | 56 | 41.4 |
| hmmt_feb_2026 | 97 | 97.1 |
| itbenchSre | 47.7 | 42.5 |
| JobBench | 54.3 | 31.3 |
| livebench_agentic_coding | 62.2 | 43.6 |
| livebench_coding | 81.5 | 74.2 |
| livebench_data_analysis | 78.7 | 71.8 |
| livebench_instruction_following | 71.4 | 74 |
| livebench_language | 85.5 | 79.7 |
| livebench_math | 84.4 | 85.3 |
| livebench_reasoning | 90.7 | 83.3 |
| MCP Atlas | 84.2 | 76.4 |
| MLS Bench Lite | 48.3 | 31.7 |
| OmniScience Accuracy | 47.6 | 31.1 |
| OmniScience Non-Hallucination | 49.1 | 74.4 |
| scicode | 58.7 | 53.5 |
| simplebench | 60.7 | 70.4 |
| simpleqa_verified | 42.7 | 58.5 |
| TauBench V3 - Banking | 46 | 11.8 |
| Terminal-Bench 2.1 | 88.3 | 75 |
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.