Mistral Small 4
Mistral Small 4
The best small open model to build a product on, and the licence has no catch.
Updated
The read
Small 4 is the model to reach for when the job is extraction, classification, routing or a first draft, and it has to run cheaply or locally. It is quick, it is Apache-licensed, it fits on modest hardware, and it behaves predictably enough to sit behind a structured output schema. Ask it to reason across many steps and it will disappoint, which is entirely fair at this size.
Where it shines
- Apache 2.0, so it can ship inside a commercial product freely
- Runs comfortably on a single consumer GPU
- Reliable structured output for its size class
Worth knowing first
- Multi-step reasoning is beyond it
- Smaller context than the large models it sits beside
Index score
78.1
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 49 of 73
- Builds
- LLMs
- Category
- Small and fast
- Output
- Apache-licensed weights, plus a cheap API
- Pricing
- Free tier
- Best for
- Cheap, local, high-volume text work
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 11Claude Haiku 5.5Claude Haiku 5.5ProvisionalA small model priced like a rounding error that scores where mid-tier models did last quarter.Small and fastPaid only87.9
- 22GPT-6 LunaGPT-6 LunaProvisionalThe cheap end of GPT-6, and the model free ChatGPT users now talk to.Small and fastPaid only84.9
- 30GPT-5.6 LunaGPT-5.6 LunaProvisionalStill cheap and still capable, but GPT-6 Luna costs half as much.Small and fastPaid only83.0
