Google Labs, Google's experimental arm, is testing a new image generator called Whisk. This tool gives you prompts using images instead of text, allowing you to remix your photos by changing the subject, scene, and style.
Whisk uses Google's image generation model Imagen 3 to combine three images: one for the subject, one for the scene, and one for the style. For example, you can choose a photo of yourself as the subject, a futuristic landscape as the scene, and an anime style as the final look.
This model automatically generates detailed captions for images. This is used by Imagen 3 as a guide when creating a remix of your photo. You can also enter text prompts to further define the desired outcome, including a detailed description such as “Subject is riding a flying bike.”
Because Whisk only focuses on a few key features of each image, the company explains that the results may not always be what you expect. For example, the generated subjects may have different heights, weights, hairstyles, and skin colors. Google says you can view and edit the underlying prompt at any time.
This experiment is currently available only to US-based users (labs.google/whisk).