Find answers from the community

Updated 4 months ago

Any way to use the messages api of

At a glance

The post asks if there is a way to use the messages API of models with images. The community members discuss that a multi-modal language model would be required to send images. They provide some examples of using the completion API with multi-modal models, but note that this does not implement the messages API. The community members acknowledge that the messages API for images has not been implemented and suggest that it should probably be done.

Useful resources

ssickness272

Any way to use the messages api of models with images???

5 comments

LLogan M

have to use a multi-modal LLM if you want to send images

LLogan M

Plain Text

resp = multi_modal_llm.complete(
    prompt="tell me about this image",
    image_documents=[image_documents[0]],
)

A few examples
https://docs.llamaindex.ai/en/stable/examples/multi_modal/replicate_multi_modal/#init-fuyu-8b-model

https://docs.llamaindex.ai/en/stable/examples/multi_modal/ollama_cookbook/?h=multimodal

ssickness272

This does not implement the messages API. this is completion api

LLogan M

ah good point. Looks like it was never implemented for some reason

LLogan M

should probably do that

Add a reply