How to Use Google Labs’ Whisk: Create Unique AI Images in Minutes

If you’ve ever wanted to generate AI visuals with precise creative control — Whisk by Google Labs might just be your new favorite tool.
This experimental image generator combines three visual elements — a Subject, a Scene, and a Style — to remix them into something totally new.

Below is a simple step-by-step walkthrough of how to use it.

Whisk (a Google Labs experiment) generates unique AI images by combining three core visual elements: a Subject, a Scene, and a Style. To be able to use Whisk you need to be verified that you are 18 years of age, you will be asked to do this later. 

Step 1: Access Whisk

Head to 👉 Whisk on Google Labs

🔐 You’ll need a supported Google account and may be asked to verify that you’re over 18.

1. Click Next, then scroll down the Privacy Policy and click Connect.

2. Complete the age verification if prompted.

3. Click Enter Tool to open the main interface — you’ll see three boxes labeled Subject, Scene, and Style, plus a text input field underneath.

Step 2: Define Your Visual Prompts

Whisk’s creativity comes from combining three ingredients. Left bar is where you upload the reference images that the AI will remix.

A. The Subject (The Main Focus)

• Click the "Subject" box.

• Upload an image of your main focus (e.g., a person, an animal, or a specific object) from your computer.

• You will be prompted that you have rights to use the image. I’ll use an image created with Whisk here. 

B. Click the Scene box and upload a background image, like a location or setting.

For example I’ll use a beach Photo by Maarten van den Heuvel on Unsplash

https://images.unsplash.com/photo-1506953823976-52e1fdc0149a?ixlib=rb-4.1.0&ixid=M3wxMjA3fDB8MHxwaG90by1wYWdlfHx8fGVufDB8fHx8fA%3D%3D&auto=format&fit=crop&q=80&w=735

C. Style – The Aesthetic

In the Style box, define the overall mood or aesthetic.
You can either type something like:

“Dramatic lighting, high contrast, cinematic tone.”
or upload a visual example — for instance, a stylized photo by Tim Mossholder on Unsplash.

Photo by Tim Mossholder on Unsplash.

Step 3: Add Text Guidance (Optional)

In the text field below, add a short prompt to guide the AI further. This helps refine poses, actions, or moods.

Example prompt:

“A woman walking on a beach turns her head back over her right shoulder, looking towards the camera. Her expression is confident and inviting.”

You can also upload a reference image to reinforce your description.

Step 4: Generate and Review

Click Generate and wait for Whisk to process your inputs.
Within seconds, you’ll get a fresh, AI-generated image — in my case, it produced a wall painting of a woman inviting the viewer to follow her along the beach.

---

Step 5: Refine and Iterate (Optional)

If your result isn’t quite right, click Refine below the image. This opens a chat-like interface where you can type adjustments, such as:

“Change the lighting to sunset,” or “Make the background more tropical."

Click Generate again to apply your refinements.

Step 6: Download your image

When you’re satisfied, click the Download icon (the downward arrow) to save your high-resolution creation.

And that’s it — your AI-generated masterpiece is ready to share! 🎨

Final Thoughts

Whisk is a fascinating experiment for creators who love visual storytelling.

By mixing subjects, scenes, and styles, you can explore endless combinations — from realistic composites to dreamlike art.

🧩 Pro tip: Try pairing unexpected styles and scenes — like a vintage portrait set in a neon city — to really test Whisk’s creative range.

If you liked this tutorial, consider subscribing or bookmarking this blog — more creative AI guides are on the way!