Has someone been able to implement Jan into home assistant assist?
seems like it would be perfect for it.
Seems like the local api server is not available yet however for a local llm it seems to run pretty smoothly.
Hi !
I’m sorry, but I don’t know how this could be possible but I recently created an integration to HuggingChat which is an Open-Source alternative to ChatGPT.
This integration doesn’t require an addon but works with the Hugging Face servers. If you’re interested, have a look at GitHub 
The Jan API does work. For Home Assistant I had to use the OpenAI Compatible Addon in HACS. Just follow the below instructions.
- In Jan load your model into memory by starting it.
- On the API page. Set the IP address to 0.0.0.0 to listen on all adapters. Make sure you unselect the CORS option since this will be a local server.
- Start the API. If you get a pop up about your firewall make sure you allow it through.
- Inside the OpenAI Compatible Conversation Plugin in Home Assistant. When you create the service set the API key to “Null” and enter http://xxx.xxx.xxx.xxx:1337/v1 for the AI server. Replacing the xxx.xxx.xxx.xxx with your IP address of the machine running Jan.
- On the configuration page for the AI. Unselect recommended and hit submit. Now make sure you specify the exact model you are using in Jan. Go to your models page in Jan and make sure you use the ID for the model that is currently loaded.
Once you save and click submit you should be able to interact with your locally hosted AI model.
Note: If you change the actively loaded model you will need to go into the OpenAI Compatible Conversation Plugin again and change the model ID to the new active model.