Devstral on Ollama

Min VRAM: 16 GB

Pull

ollama pull devstral

Run

ollama run devstral

API

After running, Ollama serves an OpenAI-compatible API at http://localhost:11434/v1.

Model card