[Feature]: customisable context deadline for ai chat #109

Preflight checks

• I searched existing issues and discussions and did not find a duplicate.

Problem / use case

Hi, I run local AI. my hardware is pretty slow and old so prefill takes it a very long time, but it works very reliably with good quality output, it's just really slow. I don't mind it being slow, but it would seem that caddyui is not patient :D

A lot of the chat requests I send to it come back with this
openai: Post "http://:8080/v1/chat/completions": context deadline exceeded

Proposed solution

I think I read somewhere in a changelog that you have a 60 second timeout applied to your chat UI. I think that's a perfectly reasonable default, but it would be really nice to be able to override that default in the AI Assistant settings.

Personally I'd probably set it to 600 seconds for my use as I'd be quite content to type in some orders and just leave it to do its thing whilst I work on something else :)

Alternatives considered

Alternative is to switch to a smaller, lighter model but as I only have 1 inference server, I don't want to faff about switching models per use-case.

Additional context

I'm running Qwen 3.6 35B A3B on an 8GB GPU with a lot of layers offloaded to the CPU so it's good quality generation, just a bit slow.