I’m using the interactions api and gemini-3-flash-preview. I have a function that generates and image and I want to send it back to the model. This seems possible in the old api, but I can’t figure out how to do it with the interactions api. Should I just put the base64 in the result object? How do I tell it that it is an image?
Hello @glibdog ,
Thanks for reaching out. We just fixed missing documentation. Example here: https://ai.google.dev/gemini-api/docs/interactions?ua=chat#multimodal_function_results
If you run into any issue please let us know.