Mistral Small 3 (24B, January 2025) is a dense open-weight model from Mistral AI tuned for low-latency chat, coding, and instruction-following. It is designed to deliver strong performance at a fraction of the cost of larger models.
Cost rate
0.1x
Context
33K
Released
Jan 30, 2025
Input
Text
Output
Text
Support
ReasoningTool callingStructured outputs
Performance
Median latency and throughput measured across recent requests.
Throughput56 tok/s
Latency0.58s
Best at
The categories where this model ranks highest.
Mathematical
#297 路 top 73%
Medicine & Healthcare
#301 路 top 76%
Software & IT Services
#310 路 top 73%
Life, Physical, & Social Science
#316 路 top 75%
| Domain | Score | Rank |
|---|---|---|
| Software & IT Services | 1305 | #310 |
| Writing, Literature, & Language | 1248 | #321 |
| Life, Physical, & Social Science | 1294 | #316 |
| Entertainment, Sports, & Media | 1225 | #321 |
| Business, Management, & Finance | 1259 | #319 |
| Mathematical | 1270 | #297 |
| Legal & Government | 1275 | #320 |
| Medicine & Healthcare | 1286 | #301 |
Latest news about Mistral Small 3
Recent articles and community posts, linked to their original sources.

Mistral CEO Explains Why AI Models Are Becoming Defence Systems20VC with Harry Stebbings7,277

AI fixes AI in MIT study, Mistral's open model & Google Genie 4IBM Technology19,765
We're releasing the weights of Mistral Large 4Hacker News1,732441
Mistral Large 4Hacker News327105