Step 1 of 4
Check your server
Step 2 of 4
Check an owned model
No running n8n was found. If the original n8n container or network is gone, enter their exact names.
This path cannot install. Removal needs its own confirmation after Relmio re-checks its files, runtime and cache.
Step 2 of 4
Choose n8n and a model
Pick n8n, its network and one CPU model. This step only looks.
Server check
Step 3 of 4
Review the plan
Nothing changes on your server until you confirm.
- Action
- Server and n8n
- Model
- Resources
- Download and disk
- Files and endpoint
/docker/n8n-openai-oauth/local-model·http://n8n-local-model:11434/v1
Downloads contact image and model registries. Prompts stay on the selected runtime, subject to n8n workflows and logs.
Technical details
Step 4 of 4
Connect n8n
- Private base URL
http://n8n-local-model:11434/v1- Model
- API key
local-onlyAn ignored placeholder. It does not authenticate the runtime.
Use Responses API: Off
This recipe uses Chat Completions. Set it in the OpenAI Chat Model node.
Start with a short fixed prompt. Inference readiness does not set up the Assistant sandbox or tool use.
- No ChatGPT sign-in
- Existing n8n stays untouched
- No model port is published