Hermes 3 on Ollama

Min VRAM: 6 GB

Pull

ollama pull hermes3

Run

ollama run hermes3

API

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

Model card