Mistral Medium 3.5 scores 24.3 on the Respan Index (#84), ahead of DeepSeek V3.2 at 24.1 (#85). DeepSeek V3.2 leads on 2 of the 3 benchmarks both report. Mistral Medium 3.5 has the larger context window (256K tokens).
| Benchmark | DeepSeek V3.2 | Mistral Medium 3.5 |
|---|---|---|
| LMArena Elo | 1425 | 1421 |
| LMArena WebDev (LMArena) | 1362 | 1263 |
| SWE-bench Verified | 73.1% | 77.6% |
Higher is better. Each score links to where it was published.
* Estimated: no published scores in that category. How the Respan Index works