Mistral AI · Mistral AI
Mistral Small 3
SupersededCost-efficient small model at $0.10/$0.30; superseded by Mistral Small 4 (Mar 2026).
About this version
Mistral Small 3 is a cost‑efficient small language model from the Mistral family that was released in early 2025 and positioned as an affordable option for inference pricing. It was marketed with low per‑token costs of $0.10 for input and $0.30 for output, and it has since been superseded by Mistral Small 4, which arrived in March 2026, moving the model into a legacy tier status.
- Tier
- Legacy
- Released
- Jan 1, 2025
Release & change timeline
1 event- Release
Mistral Small 3 released
Cost-efficient small model at $0.10/$0.30 per 1M; became the recommended budget choice.
Report an issue with Mistral AI
Posts to your status feed
Pick the closest match below, edit the body, and post. Your report carries the #mistral tag automatically so it surfaces here + in the trending-tags rail. Reporting also follows Mistral AI so you’ll get status updates.
Part of the Mistral AI version timeline. Catch Me Up — everything that changed across AI & smart home ↗
