GLM-5.1 vs Mistral Medium 3.5
| model GLM-5.1 Zhipuai | model Mistral Medium 3.5 Mistral AI | |
|---|---|---|
| Context | 200K | 262K |
| Input $/M | $0.98 | — |
| Output $/M | $3.08 | — |
| Knowledge cutoff | — | — |
| Available via | OpenRouter | Mistral AI (La Plateforme) |
| Artificial Analysis Coding Agent Index (average pass@1) | 52.7 | — |
| SWE-Atlas Codebase QnA (pass@1) | 73.2 | — |
| SWE-bench Pro (pass@1) | 19.8 | — |
| Terminal-Bench (pass@1) | 65.1 | — |
| SWE-bench Verified (resolved) | — | 77.6 |
| τ³-Telecom (accuracy) | — | 91.4 |