I don’t understand why n8n ai assistant is only available in the Cloud version. It should be possible to have it also in the self hosted versions, maybe using your own api keys. It would be very useful.
1 Like
because you don’t pay the price for this service when you are self hosted.
1 Like
Hey @jollymask, currently we only have it on our Cloud version because we need to control the models that are being called. We sometimes use different models based on the task.
As this creates a lot of costs for us, we can currently only offer this on paid versions of n8n. I hope that makes sense.
1 Like
Ok, I see. I just thought that by connecting our own API keys, it might have been possible somehow. Thanks for the explanation.
1 Like
For self-hosted instances, why not introduce an ability for users to use their own AI services that can be plugged in via environmental variables?
1 Like