| Benchmark | Opus 5 | Qwen3.7 Max |
|---|---|---|
| AA Agentic Index | 59.2 | 30.9 |
| AA Intelligence | 63.1 | 46.7 |
| AA-LCR | 78.7 | 74.7 |
| AA-Omniscience | 37.1 | 13.5 |
| arena_elo | 1487 | 1474 |
| arena_text_factuality | 1481 | 1479 |
| Artificial Analysis Coding Index | 78 | 66 |
| coding_arena_elo | 1662 | 1517 |
| critpt | 29.1 | 13.4 |
| DeepSWE 1.1 | 68.8 | 21.6 |
| Finance Agent v2 | 58.6 | 48.4 |
| gdpval | 67.2 | 38.5 |
| GPQA Diamond | 93.7 | 92.4 |
| HealthBench | 67.1 | 54.5 |
| HLE | 64.7 | 41.4 |
| HLE (with tools) | 64.7 | 53.5 |
| livebench_agentic_coding | 65.2 | 43.6 |
| livebench_coding | 81.5 | 74.2 |
| livebench_data_analysis | 74.5 | 71.8 |
| livebench_instruction_following | 63.8 | 74 |
| livebench_language | 88.7 | 79.7 |
| livebench_math | 95.7 | 85.3 |
| livebench_reasoning | 91.2 | 83.3 |
| MCP Atlas | 85.8 | 76.4 |
| OmniScience Accuracy | 60.9 | 31.1 |
| OmniScience Non-Hallucination | 40.5 | 74.4 |
| scicode | 55.7 | 53.5 |
| simplebench | 80.6 | 70.4 |
| simpleqa_verified | 56.7 | 58.5 |
| SWE-bench Multilingual | 89.5 | 78.3 |
| SWE-bench Pro | 79.2 | 60.6 |
| SWE-bench Verified | 96 | 80.4 |
| TauBench V3 - Banking | 44.7 | 11.8 |
| Terminal-Bench 2.1 | 89.1 | 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.