Mistral Small is Mistral AI's efficient model line; the current Mistral Small 4 combines chat, reasoning, coding and image input in one open model.
In plain words
Mistral Small aims for strong results at low cost and speed. Mistral Small 3, in January 2025, was a 24-billion-parameter model under the Apache 2.0 licence that rivalled models three times its size. Mistral Small 4, released in March 2026, is a mixture-of-experts model with 119 billion parameters in total but only about 6 billion active per token. It merges Mistral's separate reasoning, image and coding models into one, with adjustable reasoning effort.
A quick example
Imagine an app that sometimes needs a quick reply and sometimes a careful step-by-step answer about a photo. With Small 4 it can call the same model each time and just turn the reasoning effort up or down.
Where you'll meet it
- Mistral's API.
- Self-hosted and fine-tuned deployments.
Related
Sources
- DocsMistral Small 3, Mistral AI · read 29 Sept 2026
- DocsIntroducing Mistral Small 4, Mistral AI · read 12 Oct 2026
- DocsModels overview, Mistral AI docs · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 12 Oct 2026. Spotted an error? Report it.