GPT-OSS on Ollama

Min VRAM: 16 GB

Pull

ollama pull gpt-oss

Run

ollama run gpt-oss

API

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

Model card