search

BerriAI / litellm

Python SDK, Proxy Server to call 100+ LLM APIs using the OpenAI format - [Bedrock, Azure, OpenAI, VertexAI, Cohere, Anthropic, Sagemaker, HuggingFace, Replicate, Groq]
https://docs.litellm.ai/docs/
Other
11.65k stars 1.34k forks source link

[Feature]: Support for Reka AI #3607

Open InnerLive opened 3 months ago

InnerLive commented 3 months ago

The Feature

It would be great if you would support the multimodal models of Reka.ai.

The models understand text, images, video and audio.

The documentation is here: https://docs.reka.ai/

Motivation, pitch

I would like to be able to use the models in Open WebUI.

Twitter / LinkedIn details

No response

krrishdholakia commented 3 months ago

Hey @InnerLive any specific models / queries you're trying to run? Will help for testing

InnerLive commented 3 months ago

I hope it works well with audio files and video files. As a competitor to Google Gemini Pro. I don't have any concrete use cases, I'm more of a hobby researcher when it comes to AIs. If it helps, I'll find some files and give you some prompts.

InnerLive commented 3 months ago

You could use the videos and audios from here: https://pixabay.com/sound-effects/search/nature/ https://pixabay.com/videos/

As a prompt, then something like: What do you hear in the file? What do you see in the video? What color is [X]?

BabyChouSr commented 1 month ago

+1 to this request! Would be great to have support for Reka AI.

infwinston commented 1 month ago

Strong +1. Reka has been offering great language + vision models. Would love to see litellm supports it.