| Benchmark | Kimi K2.6 | Qwen3.6 27B |
|---|---|---|
| AA Agentic Index | 58.7 | 27.5 |
| AA Intelligence | 45.1 | 37.7 |
| AA-LCR | 76.7 | 73.3 |
| AA-Omniscience | 6.4 | -20 |
| aime_2026 | 96.4 | 94.1 |
| Artificial Analysis Coding Index | 61.8 | 53.7 |
| baby_vision | 68.5 | 28.9 |
| charxiv_rq | 86.7 | 78.4 |
| Claw Eval (pass@3) | 80.9 | 60.6 |
| critpt | 8 | 1.1 |
| gdpval | 41.4 | 32 |
| GPQA Diamond | 91.1 | 87.8 |
| HLE | 37.5 | 24 |
| hmmt_feb_2026 | 94.7 | 84.3 |
| IFBench | 76 | 69.1 |
| imo_answer_bench | 86 | 80.8 |
| livebench_agentic_coding | 46.9 | 39.3 |
| livebench_coding | 78.6 | 71.8 |
| livebench_data_analysis | 65.1 | 70.4 |
| livebench_instruction_following | 64.4 | 53.2 |
| livebench_language | 75.1 | 63.3 |
| livebench_math | 84.3 | 79.9 |
| livebench_reasoning | 79.4 | 70.3 |
| LiveCodeBench v6 | 89.6 | 83.9 |
| mathvision | 93.2 | 85.1 |
| MMLU-Pro | 87.1 | 86.2 |
| MMMU-Pro | 80.1 | 75.8 |
| OmniScience Accuracy | 32.8 | 19.6 |
| OmniScience Non-Hallucination | 60.7 | 50.7 |
| OSWorld-Verified | 73.1 | 63.9 |
| scicode | 53.5 | 39.8 |
| SWE-bench Multilingual | 76.7 | 71.3 |
| SWE-bench Pro | 58.6 | 53.5 |
| SWE-bench Verified | 80.2 | 77.2 |
| TauBench V3 - Banking | 23.3 | 16.7 |
| Terminal-Bench 2.0 | 66.7 | 59.3 |
| Terminal-Bench 2.1 | 65.9 | 60.7 |
| Terminal-Bench Hard | 43.9 | 34.8 |
| τ²-Bench Telecom (AA run) | 95.9 | 94.2 |
| τ³-Bench | 20.6 | 15.3 |
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.