Mistral AI model product
1 (2503), with added reasoning capabilities, undergoing SFT from Magistral Medium traces and RL on top, it's a small, efficient reasoning model with 24B parameters.
Updated Aug 12, 2026. Default version: Magistral Small 2506
Technical details for the model's default version.
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Benchmark scores for Magistral Small 2506.
| AIME 2024 | 70.7% | 40 | 53 | 25.0% | C | |
| LiveCodeBench | 51.3% | 43 | 73 | 41.7% | C | |
| AIME 2025 | 62.8% | 96 | 114 | 15.9% | C | |
| GPQA | 68.2% | 137 | 234 | 41.6% | C |
Preference and agent-evaluation results for the default version.
| vision | overall | 100 | 1142.7 | 10,988 | N/A | |
| vision style control | overall | 103 | 1140.7 | 10,988 | N/A | |
| text style control | overall | 183 | 1357.1 | 17,774 | N/A | |
| text | overall | 189 | 1338.8 | 17,774 | N/A |
Provider-specific output speed and catalog latency for Magistral Small 2506. Runtime does not affect the capability score.
No provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Official vendor API pricing appears first, followed by individual provider offers.
| Nvidia | mistralai/magistral-small-2506 | global | N/A | N/A | 32.8K | |
| IO.NET | mistralai/Magistral-Small-2506 | global | $0.50 | $1.5 | 128K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Available versions of this model. The score column identifies the version used in the overall ranking.
| Magistral Small 2506 | 19.8 | 24B | 131.1K | 8.2K | Yes | Apache 2.0 |
Key information about Magistral Small and its available data.
1 (2503), with added reasoning capabilities, undergoing SFT from Magistral Medium traces and RL on top, it's a small, efficient reasoning model with 24B parameters. Magistral Small can be deployed locally, fitting within a single RTX 4090 or a 32GB RAM MacBook once quantized.
Data as of 2026-08-11.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
Common questions about Magistral Small.
Magistral Small's default version was released on Jun 10, 2025.
No official standard PAYG price is currently available for Magistral Small. The lowest tracked third-party offer starts at $0.50 input and $1.5 output via IO.NET.
Magistral Small was created by Mistral AI.
The default version has a 131.1K token context window.
Yes. The default version is marked as open weight under Apache 2.0.
2 provider offerings are linked to the default version.
Nearby ranked alternatives include Nova Pro, Min istral 3, Gemini 1.5 Pro.