| Benchmark | GLM-5.1 | GPT-4o mini |
|---|---|---|
| AA Agentic Index | 66 | 1 |
| AA Intelligence | 41 | 7 |
| Artificial Analysis Coding Index | 55.8 | 11.4 |
| gdpval | 49.5 | 0 |
| GPQA Diamond | 86.8 | 42.6 |
| HLE | 52.3 | 4.2 |
| IFBench | 76.3 | 31 |
| scicode | 43.8 | 22.9 |
| simplebench | 55.1 | 10.7 |
| TauBench V3 - Banking | 13.6 | 2.9 |
| Terminal-Bench 2.1 | 63.5 | 5.6 |
| τ³-Bench | 70.6 | 2.9 |
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.