| Benchmark | Claude 3.5 Sonnet | GLM-5.2 Full Open Source |
|---|---|---|
| AA Intelligence | 10 | 52.6 |
| Artificial Analysis Coding Index | 30.2 | 68.8 |
| GPQA Diamond | 67.2 | 91.2 |
| HLE | 3.9 | 54.7 |
| scicode | 36.6 | 50.5 |
| simplebench | 41.4 | 58.8 |
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.