Mistral Small 4
Hybrid open model unifying instruct, reasoning, and coding in a single efficient package.
Our editorial assessment, based on the public sources listed below. Last reviewed .
- Overall rank
- #43
- Tier
- A
- Score
- 80.0
- Main category
- Overall
Strengths
- Single model covers instruct, reasoning, and coding modes
- Open weights with efficient footprint
- Recent v26.03 release
Considerations
- Raw capability trails larger frontier models
- Limited independent benchmark coverage so far
Price
Free (open weights); paid API available
Shown for reference — price is not part of the score.
Score breakdown
- Performance
- 80.0
- 55% of score
- Reputation
- 77.0
- 35% of score
- Recency
- 90.0
- 10% of score
The overall score combines these sub-scores using the weights shown. How we score →
Verification status
Each field carries its own re-verification date. Amber means older than 90 days — treat those figures as stale until re-checked.
Checked against the sources listed below.
Sources
- Mistral Models Overview — Mistral Small 4 (opens in a new tab)Accessed:
- Mistral Docs — Models (opens in a new tab)Accessed:
- Mistral Docs — Open Weight Models (opens in a new tab)Accessed:
- Mistral — Models from cloud to edge (opens in a new tab)Accessed:
- Mistral Docs (opens in a new tab)Accessed:
Show 3 more sources
Similar models
- Mistral Large 3Mistral AI#41 · Tier A · Score 80.9Compare
- MiniMax M3MiniMax#42 · Tier A · Score 80.0Compare
- GLM-5.2Z.ai#45 · Tier A · Score 79.9Compare
- GPT-5.4OpenAI#46 · Tier A · Score 79.6Compare