Use ChatGPT if Ollama is unavailable?

[tulip summit]

Hi,
I'm running Ollama on my gaming PC which is not on 24/7
Is there a way to fallback to ChatGPT if my Ollama server is unavailable?

[deft tusk]

You can try using Fallback Conversation agent integration from HACS. It allows picking main and fallback agents.

[tulip summit]

Awesome, thank you!

[tulip summit]

Actually, I've just looked it up and it's unfortunately not what I'm looking for 😅
This is basically what was introduced in 2024.12 - it falls back to a LLM if the query is too complex for the local conversation agent

[deft tusk]

But you can choose any agent there, innit?

Just checked, I can choose llama as first option, and OpenAI as second one.

Then just use this agent without "prefer local" thing, and that should work

[tulip summit]

Ah, sorry! The github description made it seem like it was only local fallback to LLM

Thanks! 🙂

[hasty perch]

You can make a ping integration to the ollama. and make automations to switch the pipeline of the assists online/offline