qwen3-coder does not support thinking

[Retch]

I wanted to use continue with qwen3-coder in agent mode but i can't get it to work because an error message appears. The qwen3 model is working but i want to use qwen3-coder (as recommended in the documentation).

name: Local Config
version: 1.1.0
schema: v1
models:
  - name: Nomic Embed
    provider: ollama
    model: nomic-embed-text:latest
    roles:
      - embed
  - name: qwen3-coder 30b
    provider: ollama
    apiBase: http://x.x.x.x:x
    model: qwen3-coder:30b
    roles:
      - apply
      - chat
      - edit
    capabilities:
      - tool_use

[uinstinct]

Tested it right now and it seems to be working fine.
Are you running qwen3-coder:30b with a thinking flag or have set its thinking to true by any chance?
Can you try stopping and rerunning the ollama instance? (go to the extension and start chatting. it will let continue invoke qwen3-coder:30b with the right parameters)

[Retch]

I already restarted the OLLAMA instance and set it up fresh. I don't get why continue requires the Model to think when qwen3-coder does not support thinking (stated in the ollama docs. But qwen3-coder is recommended in the continue docs.

[uinstinct]

The extension should not force thinking for qwen3-coder. Can you please record a video of the config, agent chat and the error popup in action? This will help get a better view and understanding for debugging.