To use Gemini on playground

[yu-tarrrr]

I'm using langfuse on cloud, just for my hobby.
And I'm trying to write some promt via playground, but it seems langfuse playground doesn't support Gemini now.
Is it correct? And if so, would you please tell me any workaround.
Thank you.

[dosubot[bot]]

Hey @yu-tarrrr 👋, I'm here to help you out with any issues you're facing on the playground. I can assist with bug fixes, answer questions, and guide you on becoming a contributor while you wait for a human maintainer.

The Langfuse playground currently supports all major models from OpenAI and Anthropic, in addition to Gemini. The supported models include various versions of OpenAI's GPT-4 and GPT-3.5, as well as Anthropic's Claude models. Additionally, Langfuse now supports Amazon Bedrock for LLM Playground and Evaluations ^[1]. If you're having trouble using Gemini, it might be worth checking if there are any specific configuration issues or updates needed on your end.

To continue talking to Dosu, mention @dosu.

[jannikmaierhoefer]

Hi @yu-tarrrr, you can use LiteLLM to call Gemini models in the Langfuse Playground. LiteLLM should be fully supported if you use it via the OpenAI connector in Langfuse (both evals and playground) as LiteLLM exposes an openapi-api.

[yu-tarrrr]

@jannikmaierhoefer
Thank you for your support.
I didn't know LiteLLM, so I will use it.

[kikwho]

It would still be nice to support google gemini in the playground, as well as full-fledged multimodal support (videos/images would be nice for my usecase). Not everyone wants to remotely expose a LiteLLM endpoint.

[clemra]

Hey all -- your prayers have been answered: https://langfuse.com/changelog/2024-11-28-google-vertex-ai-support-playground-evals

thank you @hassiebp for shipping this!