Features
Text to image generation limited to puppy subjects, with a single prompt box and no style or model controls. Prompts are accepted in English and Chinese. Every output gets its own shareable result page, and past generations are gathered into public explore galleries that display the original prompt under each thumbnail.










