Support for Internal GENAI APIs Without Proxy for Aider Compatibility
Issue
We are currently working with internal GENAI APIs, which are not directly compatible with OpenAI or other models. To make Aider work with these APIs, we have to use a proxy to bridge the compatibility gap.
Question: Is there a better or more streamlined way to connect Aider to our internal GENAI APIs without needing a proxy? Are there any plans to support custom or internal API configurations that aren’t directly compatible with OpenAI models?
Version and model info
Aider :0.56.0
Thank you for filing this issue.
As far as I know there are currently no plans to use other methods than those provided via LiteLLM (-> https://github.com/BerriAI/litellm) to communicate with LLM endpoints.
Aider can work with hundreds of providers via litellm. You'll need to use a proxy if your internal API isn't compatible with any of them.
I'm going to close this issue for now, but feel free to add a comment here and I will re-open. Or feel free to file a new issue any time.