Side-by-side comparison of Gemma 4 (Google · USA) and Mistral Small 4 (Mistral AI · France) for self-hosted deployment of the open-weight model. Gemma 4 is rated EU-ready; Mistral Small 4 is EU-ready. They part ways on training data: Gemma 4 is "Partial", Mistral Small 4 is "Undisclosed".
| Field | ||
|---|---|---|
| Summary | ||
| Verdict | EU-ready Major licence shift from Gemma 2/3: Apache 2.0 across the family. 140+ languages, multimodal (text/image/audio/video on small sizes), 128K-256K context. Strong permissive default for EU deployments that need robust multilingual support. | 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 | Apache 2.0 | Apache 2.0 |
| Commercial use | Yes | Yes |
| Training data | Partial | Undisclosed |
| Origin | USA | EU |
| Performance & pricing? | ||
| Quality index | 32/100 | 19/100 |
| Speed | — | 147 tok/s |
| Blended price | — | $0.26/M |
| Context window | — | — |
| Evidence | ||
| Sources | ||