Allow multiple openAI for settings/chat

Feature Requests
1

The idea is:

settings/chat allows only one openAI connection. Allow adding multiple with custom names.

My use case:

I’m using openAI, Nous Research subscription, and local oMLX, all which use openAI.

I think it would be beneficial to add this because:

I can only use one provider at a time, and openAI is most common protocol out there.

Any resources to support this?

Are you willing to work on this?

2

Welcome @calmsofa4266!

Solid use case - running OpenAI, Nous Research, and a local oMLX stack in parallel is exactly the kind of multi-provider setup that settings/chat doesn’t support today. As a workaround for now, note that within actual workflows each OpenAI Chat Model node lets you select a different credential, so you can already switch providers per-workflow or even per-node. The gap is specifically in the settings/chat assistant UI, which is limited to a single credential. Voted - this would make n8n much more useful for testing and comparing providers side by side.

3

@calmsofa4266 Can you provide more information on why you would want to do this and how you would want it to work?

4

I’d like to be able to alternate between few subscriptions (OpenAI, nous research), and local, all which use OpenAI protocol.

How it should work: option that says “custom OpenAI endpoint”, and ideally be able to add several. This already works elsewhere in n8, where we can add as many OpenAI compatible endpoints as needed.