This tweet describes a process where four images generated by Nano Banana Pro are used as input for a video generated by Kling 3.0 Omni. The prompt, which was designed by GPT-5 and used for both image and video generation, describes a cinematic close-up of a person enjoying making and tasting sweets.
EnglishA cinematic close-up of a person enjoying making and tasting sweets. Depicting everything from the serious expression while cracking eggs and mixing cream, to the adorable gesture of delight upon completion, and finally, the full smile of satisfaction after taking a bite. The scene is captured with an intimate distance that conveys their breathing and in quiet lighting.
Click for full size · wheel to zoom · drag to pan