🇫🇷 France🧠 GeneralistReleased January 2025
Mistral Small 3
Mistral AI
Open-sourceFreeSelf-hostRGPD
Verdict
A solid pick for self-hosting and full data control.
Overview
Mistral Small 3 (24B) is under Apache 2 license, allowing free deployment on your own servers (even a decent GPU is enough). Ideal for on-premise, sovereignty-constrained apps, or edge computing.
Skill profile
Arena Elo1235
Verified benchmarks
MMLU75%
HumanEval80%
What do these scores mean?
MMLU75%very good
General knowledge across dozens of academic subjects.
HumanEval80%excellent
Python code generation from specifications.
Strengths
- Strong French quality
- Has a free tier
- Open-source and self-hostable
- Competitive input pricing
Who is it for
A good fit if…
- you want to control cost or self-host
- you work in French
- you have GDPR constraints
Skip it if…
Ideal use cases
- Lightweight self-hosting
- Mobile apps
- Volume
- On-premise enterprise
Access & availability
Paid API per tokenFree tier availableSelf-hostable (open weights)
Key specifications
Context
32K
Input price
$0.20 $/M
Output price
$0.60 $/M
Speed
150 tok/s
Price not auditedBenchmarks not audited
Estimate your monthly cost
Per-token API pricing$10/ month with Mistral Small 3
For the same usage
- Gemini 3 Flash$36+253%
- Gemini 3.1 Pro$145+1314%
- GPT-5.5$363+3434%
Indicative estimate based on standard API rates (excluding caching, batch and volume discounts). Always check the official pricing before committing.
Privacy
GDPR-compliantHosting : EU/selfAnonymizable data
Advanced data · for expertsArchitecture, modalities, detailed cost, full benchmarks▾
Advanced data · for experts
Architecture, modalities, detailed cost, full benchmarks
Architecture
Dense
Size
24B
Cutoff date
Sep 2024
Inputs
text
Outputs
text, code
License
apache-2
Hosting
EU/self
Hallucination score
3/5
Estimated cost (API)
Typical exchange (~3k in / 1k out)
≈ $0.0012
1M in + 1M out
≈ $0.80
All benchmarks
| Arena Elo | 1235 |
| MMLU | 75.0% |
| GPQA | — |
| HumanEval | 80.0% |
| SWE-Bench | — |
| MATH | — |