Side-by-side comparison of Mistral Large 2 (Mistral AI · France) and Mistral Small 4 (Mistral AI · France) for self-hosted deployment of the open-weight model. Mistral Large 2 is rated conditional; Mistral Small 4 is EU-ready. They part ways on licence: Mistral Large 2 is "MRL (research)", Mistral Small 4 is "Apache 2.0".
| Field | ||
|---|---|---|
| Summary | ||
| Verdict | Conditional Frontier 123B dense model from an EU vendor. Open weights under Mistral Research License — non-commercial by default. Commercial deployment requires a separate paid licence from Mistral AI. | EU-ready Unified model folding Instruct, reasoning (Magistral) and code (Devstral) into a single 119B MoE under Apache 2.0. 6.5B active params, 256K context, 24 languages, toggleable reasoning effort. Strongest permissive EU option at this scale. |
| Last reviewed | 2026-04-15 | 2026-04-15 |
| Open-weight | ||
| Licence | MRL (research) | Apache 2.0 |
| Commercial use | Paid licence req. | Yes |
| Training data | Undisclosed | Undisclosed |
| Origin | EU | EU |
| Performance & pricing? | ||
| Quality index | 15/100 | 19/100 |
| Speed | 38 tok/s | 147 tok/s |
| Blended price | $3.00/M | $0.26/M |
| Context window | — | — |
| Evidence | ||
| Sources | ||
No overlapping sources between the two entries.