zhu327 / gemini-openai-proxy

A proxy for converting the OpenAI API protocol to the Google Gemini Pro protocol.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

[Note] Client Compatibility List and Known Issues

zhu327 opened this issue · comments

Perfectly Compatible:

Known Issues:

  • There is a bug in the Google Gemini Pro SDK, which may lead to unexpected interruptions during streaming output.
  • If your prompt contains keywords like "openai," "chatgpt," "gemini," etc., it might be rejected by Gemini Pro for a response.

Feel free to contribute to the compatibility list, allowing us to continuously enhance the user experience of our project. Thank you for your support and feedback!

https://ai.google.dev/tutorials/go_quickstart?hl=en#multi-turn-conversations-chat

Note: The gemini-pro-vision model (for text-and-image input) is not yet optimized for multi-turn conversations. Make sure to use gemini-pro and text-only input for chat use cases.

Due to Google Gemini Pro Vision's lack of support for chat context, context will be lost when handling gpt-4-vision-preview conversations.

I use this project at https://harpa.ai.
Because the author of the extension does not want to provide direct calling of the Gemini Pro API, only requiring users to utilize Gemini Pro through OpenRouter.

Seems to be working with shell gpt TheR1D/shell_gpt#514