LLM
Devstral Small 2
Mistral AI
Updated May 22, 2026 · Benchmarks via Artificial Analysis, specs & pricing via OpenRouter · Methodology · Report an error
Specifications
- Type
- LLM
- Released
- December 9, 2025
- Output speed
- 62 tok/s
- Latency (TTFT)
- 0.75s
- API pricing
- $0.00 in · $0.00 out / 1M tokens
Benchmarks
Reasoning
| GPQA Diamond | 53.2 |
Coding
| LiveCodeBench | 34.8 |
Math
| AIME 2025 | 34.3 |
General
| MMLU-Pro | 67.8 |
| Humanity’s Last Exam | 3.4 |