MultimodalOpen weights
Mistral Small 4
Mistral AI
Updated May 21, 2026 · Benchmarks via Artificial Analysis, specs & pricing via OpenRouter · Methodology · Report an error
Specifications
- Type
- Multimodal
- Access
- Open weights
- Released
- March 16, 2026
- Context window
- 262,144 tokens
- Output speed
- 145 tok/s
- Latency (TTFT)
- 0.51s
- Input
- Text, Image
- Output
- Text
- API pricing
- $0.15 in · $0.60 out / 1M tokens
Capabilities
Function callingStructured output
Benchmarks
Reasoning
| GPQA Diamond | 76.9 |