I recently started testing out the Gemini Pro models since Google is currently offering free access to their API and I ended up creating this AI Stream Interact library which allows for interacting with the model over a video stream out of your web cam. Curious to hear thoughts and gather feedback from those who may be interested.