Enhancing LiteLLM Support in LobeChat

[aur3l14no]

Hi team! 👋

Following up on @arvinxx's previous comment about prioritizing LiteLLM as first-class citizen in LobeChat, I wanted to propose a wishlist to strengthen this integration.

Feature Wishlist

1. Fetch model info using LiteLLM's /model/info endpoint
2. Dedicated Provider Implementation
   Consider adding LiteLLM as a separate provider rather than relying solely on OpenAI compatibility patches (reference). This could:
   • Prevent conflicts with native OpenAI configurations (eg. model preset)
   • Better handle LiteLLM-specific parameters/features
   • Simplify maintenance long-term

Thanks a lot for building this amazing app. I'm excited to see where this integration could go!

[arvinxx]

Really good idea!