A multi-step conversational prompt used to refine an image generation request for a VTuber gaming stream thumbnail. The initial request asks for a character (based on an uploaded image) to be shown intensely focused on a game controller, maintaining their style and outfit. The subsequent steps refine the character's appearance, requesting them to be more delicate, cuter, with a smaller bust size and long sleeves, and finally correcting a text error in the generated image.
Maintain the character's style and outfit, change the scene to them intensely gripping a game controller, and create a thumbnail image in the style of a fictional VTuber's game stream. They look a bit too robust, so please make them more delicate and cute. Also, the bust size should be more modest, and the sleeves should be long sleeves. Fix the collapsed 'Ge' in the top left corner.
Generate asks for your photo, then opens the editor with the prompt already loaded.
Everything after the generation, on the same site.
Same model, same dataset β GPT Image 1.5 prompts people have shared.