GLM-5.3 vs Mistral Medium 3.5
| model GLM-5.3 Zhipu AI | model Mistral Medium 3.5 Mistral AI | |
|---|---|---|
| Context | 1M | 262K |
| Input $/M | — | — |
| Output $/M | — | — |
| Knowledge cutoff | — | — |
| Available via | weights only (no API yet) | Mistral AI (La Plateforme) |
| SWE-bench Verified (resolved) | — | 77.6 |
| τ³-Telecom (accuracy) | — | 91.4 |