| Benchmark | GLM-5.1 | Qwen3.7 Plus Preview |
|---|---|---|
| AA Agentic Index | 66 | 20.8 |
| AA Intelligence | 41 | 39.4 |
| AA-LCR | 68 | 69 |
| AA-Omniscience | 0.8 | 2.4 |
| Apex (Pass@1) | 11.5 | 22.7 |
| arena_elo | 1468 | 1458 |
| arena_text_factuality | 1458 | 1454 |
| Artificial Analysis Coding Index | 55.8 | 55.9 |
| critpt | 4.6 | 9.1 |
| Finance Agent v2 | 44.8 | 38.2 |
| gdpval | 49.5 | 22.2 |
| GPQA Diamond | 86.8 | 90.3 |
| HLE | 52.3 | 35.6 |
| hmmt_feb_2026 | 89.4 | 92.9 |
| IFBench | 76.3 | 79.1 |
| imo_answer_bench | 83.8 | 86 |
| MCP Atlas | 75.6 | 73.2 |
| MMLU-Pro | 86 | 88.5 |
| nl2repo | 42.7 | 41.1 |
| OmniScience Accuracy | 25.2 | 22.5 |
| OmniScience Non-Hallucination | 70.1 | 74.5 |
| scicode | 43.8 | 51.3 |
| SWE-bench Multilingual | 73.3 | 75.8 |
| SWE-bench Pro | 58.4 | 57.6 |
| TauBench V3 - Banking | 13.6 | 17.9 |
| Terminal-Bench 2.0 | 69 | 70.3 |
| Terminal-Bench 2.1 | 63.5 | 61 |
| Terminal-Bench Hard | 43.2 | 47 |
| τ²-Bench Telecom (AA run) | 97.7 | 93 |
| τ³-Bench | 70.6 | 17.9 |
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.