I’m trying to generate images in Google AI Studio, but no matter which model I use (Pro 1.5, Flash 1.5, Flash 2.0, etc.) I get the answer:
As a text-based AI, I cannot physically draw an image. However, I can create a visual description using text characters to approximate what the image would look like.
Why? I can generate images from the Gemini iOS app. Is it due to the fact that I’m on the free tier?
Hi @Sergio_Rossi, Welcome to the forum.
Gemini 1.5 version supports multimodal input but only generates text responses. On the other hand, Gemini 2.0 handles both multimodal input and output.
However, the ability to generate images and audio is currently available only through a private experimental release.
Thanks
The Gemini iOS app and https://gemini.google.com/app are consumer applications. AI Studio is a development environment.
For the consumer application, there are backend connections to other apps on the Google cloud, including the image generator.
Hope that explains why these environments are different: they are intended for different audiences who will normally have a different goal in mind when using each.
Thank you a lot @GUNAND_MAYANGLAMBAM! I guess it’s not possible to become a beta-tester fro the experimental release. Do you have an ETA for when image and audio generation will be publicly available?