Hi all,
I’m running n8n locally on Olares OS and I’m hitting a wall with the Ollama integration in an AI Agent workflow.
Setup:
- n8n running on Olares OS
- Ollama Chat Model node connected to an AI Agent node (with Simple Memory)
- Workflow:
When chat message received→AI Agent→Ollama Chat Model
The problem:
Every time I submit a message in the chat, I get Error: Failed to receive response. The Logs panel stays empty — no error details there either.
What I’ve already checked:
- The connection to Ollama works — in the Ollama Chat Model node I can see and select the available models, so the Base URL is reachable.
- “Allowed HTTP Request Domains” is set to
All. - The credential is configured (Base URL + API key).
So the model list loads correctly, but the actual chat generation always fails with “no response.”
Questions:
- Has anyone run into this on Olares specifically?
- Could this be a timeout / resource issue with CPU inference, or something about how Olares networks the n8n and Ollama containers?
- Is there a recommended Base URL / config for reaching Ollama from the n8n container on Olares?
Any pointers appreciated. Happy to share more config details. Thanks!
