Gemini image generation limitations
Summary
Technical documentation detailing the language support, input constraints, and behavioral limitations of Gemini 2.5 Flash Image and Gemini 3 Pro Image generation models.
Key quotes
Image generation doesn't support audio or video inputs.
If a prompt is potentially unsafe, the model might not process the request and returns a response indicating that it can't create unsafe images.
This documentation outlines specific constraints for Vertex AI image generation, including language optimization and maximum image input limits. It also describes failure modes related to prompt ambiguity and safety filters.