With Nano Banana it was possible to generate an image and simultaneously output a text alongside it.
A prompt like:
”Return an image of a space shuttle launch then describe what you would expect to see in the image. Return an image and the text description.”
Would return the image and a text. With Nano Banana Pro this is no longer possible. I am guessing this is due to the thought of chain?
Maybe there is a trick but I was not able to get both image and text from a single api call. Obviously i could send the image back to gemini and request a description but i was hoping to skip this additional step.
If it is intended that output modalities “image + text“ should be able to return both in the same prompt consider this a bug report. If it is not currently within the scope of the models capabilities this is more of a feature request.
-Thanks for the awesome model