Local Model Backend Not Running | Generated by AI

Home 2026.05

The error is clear: the local model backend isn’t running.

The system tried to call openai/gpt-oss-120b at http://localhost:1234/api/v1/chat and got Connection refused — there’s no model server listening on port 1234.

Per the README:

A local OpenAI-compatible chat backend reachable at http://localhost:1234/api/v1/chat

You need to start a local inference server (e.g. LM Studio, Ollama, llama.cpp server, vLLM) that:

Quick check — do you have LM Studio or Ollama installed? If so, start it and load a model, then the session creation will work.


Back Donate