mistralai/Devstral-2-123B-Instruct-2512
| benchmark | category | score | context | conf | evidence |
|---|---|---|---|---|---|
| SWE-bench Verified | swe | 72.2 | low | | **Devstral 2** | 123 | 72.2% | 61.3% | 32.6% | | |
| Terminal-Bench | agentic-coding | 32.6 | low | | **Devstral 2** | 123 | 72.2% | 61.3% | 32.6% | |
none scraped.