Mistral Mistral Medium 3.5 | Moonshot AI Kimi K2 (0905) | |
|---|---|---|
| Overview | ||
| Company | Mistral | Moonshot AI |
| Release date | Apr 29 2026 | Sep 5 2025 |
| Model type | — | — |
| Open source | Yes | Yes |
| Specifications | ||
Parameters | 128B | 1T |
Context window | 256k | 256k |
| Benchmarks | ||
Coding SWE-Bench VerifiedReal coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 77.6%Best | 65.8% |
Science GPQA DiamondGraduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | — | 75.1% |
| Timeline | ||
| Release gap | Kimi K2 (0905) shipped 236 days before Mistral Medium 3.5 | |
Mistral Medium 3.5 leads Kimi K2 (0905) on 1 of the 1 benchmark they both report (SWE-Bench Verified). Kimi K2 (0905) shipped 236 days before Mistral Medium 3.5, so benchmark comparisons should account for the intervening progress.
Mistral Medium 3.5 has 128B parameters, while Kimi K2 (0905) has 1T. Context windows are 256k (Mistral Medium 3.5) vs 256k (Kimi K2 (0905)).
On SWE-Bench Verified, Mistral Medium 3.5 leads at 77.6% vs Kimi K2 (0905) at 65.8%.