Mistral Medium 3.5
128B · SWE-V 77.6 beats Devstral 2 + Qwen3.5; τ³-Telecom 91.4
- Vendor
- Mistral
- License
- open weights
- Released
- 2026-04-30
- Parameters
- 128B
- Context window
- 128k tokens
Capabilities
Mistral Medium 3.5 supports tool use and extended reasoning, with a 128k-token context window.
- Vision — image input (not confirmed)
- Audio — audio input (not confirmed)
- Tools — function calling / tool use
- Reasoning — extended thinking
- Computer use — desktop / browser control (not confirmed)
Benchmarks
| SWE-Bench Verified | 77.6% |
|---|---|
| SWE-Bench Pro | Not published |
| Terminal-Bench | Not published |
| Reasoning composite | 80 / 100 |
The reasoning composite is a directional 0–100 score aggregated across public reasoning evals — comparable within this matrix, not an official vendor number.
Pricing
$0.6/M input tokens.