Support for multi-modal models in Playground?

[leostudiooo]

Select Topic Area

Question

Body

It gets me thinking when I head to multi-modal models such as OpenAI's GPT-4o or Phi-3.5 vision instruct, where there aren't any "add attachment" icons or buttons in the Playground input box.

Is this going to be a feature supported in future versions of Models Playground? Or if it's that I just haven't discovered the option yet?

[matthewisabel]

Multi-modal input is coming soon!