OPEN Web UI Compatibility

[dpnmw]

Is there any consideration to allowing opensource chat clients the ability to use this mcp?

Could i do something like mcpo --port 8000 -- npx n8n-mcp and pipe to an openweb ui instance?

[czlonkowski]

@dpnmw give it a try and share how your experiments went - did it work?

[xcalibur2k]

Yeah I'm curious about this too. Work for a small business and we are trying to run everything local until we can justify paying for cloud services.
LM Studio compatibility would be fantastic too using OpenAI endpoints

[jayc-codes]

Wow wish I saw this post before I spent the whole day figuring it out...

I was sick of the claude limits (even with Pro) and I didn't want to use windsurf / cursor / copilot as ultimately they all have limits in some way.

So I just managed to get openwebui working with deepseek-reasoner in almost the same way:
mcpo --port 8001 -- npx n8n-mcp

The OP was bang on... it works a charm!

Just get mcpo running and configured to your n8n API .. with some export like :
export N8N_API_URL='https://ngrokurl'
export N8N_API_KEY='your_n8n_api_key_here'

add your MCPO url (http://host.docker.internal:8001) in openwebui > admin settings > tools

then configured openwebui for the extra host (I used docker):
docker run --add-host host.docker.internal:host-gateway
--name open-webui
-p 3001:8080
-v open-webui:/app/backend/data
-e OLLAMA_BASE_URL=http://localhost:11434
--restart unless-stopped
ghcr.io/open-webui/open-webui

openwebui > admin settings > model > find your model and add the MCPO Tool and System Prompt was the original instructions

Now while it works and it can upload workflows, it doesn't work as well as Claude.

Also deepseek-chat is working a little better than deepseek-reasoner

If anyone has any tips or ideas to optimise this further it would be great to know..