Mistral Medium 3.5 128B

About this model

Mistral Medium 3.5 128B is a dense 128‑billion‑parameter Mistral‑3 model with a 256k context window. It supports multimodal input (text and images) and offers instruction‑following, reasoning, and coding capabilities with native function calling and JSON output. In benchmarks it achieves 91.4% on τ³‑Telecom and 77.6% on SWE‑Bench Verified, demonstrating strong agentic performance. The model is released under a Modified MIT License.

Technical specifications

Capabilities

Input modalities

Output modalities

Reasoning

Hybrid Default off

Knowledge horizon