Claude Opus 5 vs Mistral Medium 3.5
| model Claude Opus 5 Anthropic | model Mistral Medium 3.5 Mistral AI | |
|---|---|---|
| Context | 1M | 262K |
| Input $/M | — | — |
| Output $/M | — | — |
| Knowledge cutoff | 2026-05 | — |
| Available via | Anthropic API | Mistral AI (La Plateforme) |
| SWE-bench Verified (resolved) | 96 | 77.6 |
| SWE-bench Pro | 79.2% | — |
| SWE-bench Multilingual | 89.5% | — |
| SWE-bench Multimodal | 59.4% | — |
| DeepSWE | 68.8% | — |
| FrontierCode (mean@5) | 53.4 | — |
| τ³-Telecom (accuracy) | — | 91.4 |