| Benchmark | GLM-5.1 | o1 |
|---|---|---|
| AA Agentic Index | 66 | 31.1 |
| AA Intelligence | 41 | 24 |
| AA-LCR | 68 | 63.3 |
| AA-Omniscience | 0.8 | -10.6 |
| Artificial Analysis Coding Index | 55.8 | 39.7 |
| critpt | 4.6 | 0.3 |
| gdpval | 49.5 | 11.5 |
| GPQA Diamond | 86.8 | 75.7 |
| HLE | 52.3 | 7.7 |
| IFBench | 76.3 | 70.3 |
| livebench | 70.2 | 52.3 |
| OmniScience Accuracy | 25.2 | 34.7 |
| OmniScience Non-Hallucination | 70.1 | 30.7 |
| scicode | 43.8 | 35.8 |
| simplebench | 55.1 | 41.7 |
| Terminal-Bench Hard | 43.2 | 12.9 |
| τ²-Bench Telecom (AA run) | 97.7 | 62.6 |
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.