Google Unveils Whisk: An AI-Powered Image Generator for Creative Exploration

Google introduces Whisk, an AI tool that transforms images into creative prompts, ideal for brainstorming and rapid visualizations.

Google has launched a new AI tool named Whisk, designed to revolutionize the way we use images as prompts for creative projects. Unlike traditional image generators, Whisk focuses on capturing the essence of your initial image, making it perfect for brainstorming and quick visualizations rather than detailed edits.

The tool features a straightforward interface, allowing users to select from three predefined styles: sticker, enamel pin, and plushie. These options are tailored to produce outputs that align with Whisk’s current capabilities. For instance, it successfully generated a plushie version of Wilford Brimley, showcasing its potential despite some limitations.

For those seeking more control, Whisk offers an advanced editor mode. Here, users can input text or images across three categories: subject, scene, and style, along with an option to add descriptive text for final touches. However, the advanced mode’s results may not always meet expectations, as seen in attempts to create specific scenes or styles.

Google emphasizes that Whisk is best used for rapid visual exploration rather than producing final content. The tool leverages the Gemini language model to interpret the source image and generate a new image based on that interpretation, not the original image itself. This approach explains why the outputs may vary in aspects like height, weight, hairstyle, or skin tone from the source image.

Currently, Whisk is available exclusively in the US. Interested users can explore its capabilities on the Google Labs site.