Mistral: Mistral Medium 3.5
● Excellent8.3
/ 10
Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex...
Specifications
| Attribute | Value |
|---|---|
| Lab | Mistral AI |
| Tags | Intelligent Coding Agentic |
| Overall Score | 8.3/10 |
| Release Date | 2026-04 |
| Context Window | 262,144 tokens |
| Input Price / 1M | $1.50 |
| Output Price / 1M | $7.50 |
| Input Modalities | Text, Image, File |
| Output Modalities | Text |
Strengths
- Dense 128B architecture for consistent quality
- Strong instruction following
- Good agentic and coding performance
- Supports image+text inputs
Weaknesses
- Below Mistral Large 3 on complex reasoning
- Moderate context window for its class
- Higher latency than MoE alternatives
Best For
- Instruction-following tasks
- Agentic coding workflows
- Multilingual enterprise apps
Sources & Further Reading
Related Models
Scores are aggregated from public benchmarks (MMLU, HumanEval, MATH, GSM8K, LMSYS) and normalized to a 1–10 scale. Methodology →