Google Introduced “Whisk”, an artificial intelligence tool that creates images using images!

Google has introduced a new artificial intelligence tool called Whisk, designed to simplify visual content creation by reducing the reliance on text commands. This innovative model uses visual commands to generate customized visuals, setting a new standard in AI-powered design.

A New Approach to Visual Creation

During its recent announcement event, Google unveiled several AI-powered tools, including Veo 2, a video creation tool aimed at competing with OpenAI’s Sora. Among these introductions, Whisk stood out for its unique ability to create visuals using reference images instead of extensive text input.

How Whisk Works

With Whisk, users can guide the AI by providing reference images that define the main elements, details, and desired style of the visual output. There’s no restriction on using just one image—multiple images can be referenced to create a blended design that meets user preferences.

While the focus is on visual commands, users still have the option to supplement their input with text commands. This feature becomes particularly useful when specific details are hard to represent visually or when suitable reference images are unavailable.

Once the input process is complete, Whisk generates a variety of visuals that align with the provided references. Users can select their favorite version, download it, or even make further adjustments using text-based prompts if desired.

Not for Photorealism, but for Unique Design

It’s important to note that Whisk is not designed for creating photorealistic images. Instead, it targets users seeking distinctive and artistic designs. Its underlying technology is powered by Google’s latest visual rendering model, Imagen 3.

Currently, Whisk is not publicly available and is inaccessible in certain regions. Google plans to expand access in the near future. We will share updates as soon as the tool becomes widely available.

Google’s video for Whisk