GPT-5.6 Luna vs Mistral Medium 3.5
| model GPT-5.6 Luna OpenAI | model Mistral Medium 3.5 Mistral AI | |
|---|---|---|
| Context | 1.1M | 262K |
| Input $/M | — | — |
| Output $/M | — | — |
| Knowledge cutoff | 2026-02 | — |
| Available via | OpenAI API | Mistral AI (La Plateforme) |
| SWE-bench Pro | 62.7% | — |
| Terminal-Bench (success rate) | 84.7 | — |
| DeepSWE | 67.2% | — |
| GPQA Diamond (accuracy) | 92.3 | — |
| FrontierMath (accuracy) | 78.6 | — |
| BrowseComp (accuracy) | 83.3 | — |
| SWE-bench Verified (resolved) | — | 77.6 |
| τ³-Telecom (accuracy) | — | 91.4 |