Images and videos
Turn a prompt or reference image into a creation.
Image generation
Open Image generation. Describe the subject, setting, visual style, and composition. Add reference images when the selected mode supports them.
Image models include Nano Banana 2 and GPT Image 2. Choose from the aspect ratios offered in the form, such as square, landscape, portrait, or 4:3. The supported output resolution is 1K.
A useful prompt is specific: “A playful orange cat wearing headphones, studio photograph, full body, plain blue background.” Refine one or two details at a time when iterating.
Video generation
Open Video generation. Write what happens, how the subject moves, and how the camera should behave. For an image-led mode, add the required reference image before generating.
The supported Veo 3.1 Fast flow produces an 8-second, 720p video with 16:9 or 9:16 framing. Select a mode and use the settings actually available in the form. Reference-image requirements vary by mode.
For a short social clip, keep the prompt focused on one action. For example: “The cat nods to the music and raises one paw; fixed camera, soft studio lighting.”
Find the result
Generate once, then follow the task in history. You can leave the editor while processing continues and return to inspect the result. A completion notification links to the finished creation.
Review the output before sharing. To manage files or download a result, read uploads and history.