# What Happened to Google Whisk? A Whisk AI Alternative

If you searched for "Whisk AI" and landed here, you are probably looking for Google's Whisk, the Labs experiment that let you remix a subject, a scene and a style into a new image, or for something that works the same way. This post gives a short, sourced history of Google Whisk, explains how to check what Google offers today, and shows how WhiskAI, an independent Whisk-style remixer, recreates the workflow with current models. We try to be precise about what we know and what we don't.

**To be clear up front:** WhiskAI is not Google's Whisk. It is an independent service, not affiliated with, endorsed by or sponsored by Google. Whisk and Google are trademarks of Google LLC. We use "Whisk AI" only to describe the remix workflow.

## A short history of Google Whisk

Google Labs announced Whisk in December 2024 on the official Google blog ([blog.google: "Whisk: Visualize and remix ideas using images and AI"](https://blog.google/technology/google-labs/whisk/)). The idea was simple and new at the time: instead of writing a detailed prompt, you dragged in images for a subject, a scene and a style. According to Google's announcement, Gemini wrote a caption for each image behind the scenes and Imagen 3 generated the result. Google described it as a tool for fast visual exploration rather than precise editing, and noted that the result captures the "essence" of the subject rather than an exact copy.

Whisk quickly became popular for playful formats: stickers, plush toys, enamel pins, figurines and posters made from a pet photo or a sketch. That vocabulary is still what people search for when they type "whisk ai".

## Is Google Whisk still available?

We are not going to guess. Google Labs tools change often, are sometimes limited by country, and are sometimes merged into other products. You will find blog posts, videos and forum threads online claiming that Whisk has shut down, come back or moved, but those are not primary sources, and we have not seen an official Google announcement we can cite for any of them.

The reliable way to check is to go to the source:

- **labs.google** lists the experiments Google currently offers and where they are available.
- **blog.google** (the Google Labs section) is where Google announces launches, changes and retirements.
- **Flow** is Google's AI filmmaking tool, and Google has presented it as its home for creative generation with its latest models; check labs.google for whether image remixing is part of it in your country.

If Google's tool is available to you and does what you need, use it. If it isn't, or you want a few things it doesn't offer, the rest of this post is for you.

## How WhiskAI recreates the Whisk AI remix

WhiskAI keeps the part people loved, the three labelled inputs, and builds it on models you can choose yourself. In the [image studio](https://whiskai.net/image):

1. **Three slots: Subject, Scene and Style.** Fill any of them. Each slot is labelled, and the studio tells the model which image plays which role.
2. **An optional prompt.** Leave it empty for a pure image remix, or add a sentence for pose, text or composition.
3. **A model you pick.** Nano Banana Pro (Google's Gemini image model, used here through an API provider) is the default because it takes up to 8 references and keeps subjects recognisable. Nano Banana 2.1 and Nano Banana 2 are cheaper or more detailed options, and GPT Image 2 and GPT Image 2.5 Flare give a more illustrative look with strong lettering.
4. **Animate this.** Any result can be sent to the [video studio](https://whiskai.net/video) in keyframes mode as the first frame, so a sticker or a clay figure can become a short clip.

Our step-by-step guide, [how to use Whisk AI](https://whiskai.net/blog/how-to-use-whisk-ai), covers each slot in detail, and [Whisk AI style ideas](https://whiskai.net/blog/whisk-ai-style-ideas) has ten prompt formulas to start from.

## Honest comparison: Google Whisk and WhiskAI

|                     | Google Whisk (as announced in Dec 2024)  | WhiskAI                                                       |
| ------------------- | ---------------------------------------- | ------------------------------------------------------------- |
| Who runs it         | Google Labs                              | An independent team, not Google                               |
| Inputs              | Subject, scene, style images             | Subject, scene, style slots plus optional prompt              |
| How images are read | Gemini captions, then Imagen 3 generates | Images go directly to a multi-reference model                 |
| Model choice        | Fixed                                    | Nano Banana Pro, 2.1, 2, Lite; GPT Image 2, 2.5, 1.5 and more |
| Resolution          | Set by Google                            | 1K, 2K and 4K depending on the model and plan                 |
| Video               | Check Google's current offering          | Animate this → image-to-video with Veo 3.1, Seedance or Kling |
| Price               | Free Labs experiment where available     | Pay per image with credits; free signup credits to try        |
| Availability        | Limited by country; check labs.google    | Works in any modern browser                                   |

Two differences matter most in practice. First, because the reference images go straight to the model rather than through a text caption, WhiskAI remixes usually stay closer to the original subject; that is great for products and pets, but you may get less "surprise". If you want looser results, use a weaker subject image or describe the change in the prompt. Second, WhiskAI is a paid service: you see the exact credit cost before each run, and failed renders are refunded.

## What a Whisk AI remix costs here

Prices come straight from our model list. A remix is one image generation:

- **Nano Banana Pro:** 25 credits at 1K or 2K, 33 at 4K.
- **Nano Banana 2.1:** 6 credits at 1K, 9 at 2K, 13 at 4K.
- **Nano Banana 2:** 11, 17 or 25 credits at 1K, 2K or 4K.
- **GPT Image 2 / 2.5 Flare:** 9, 14 or 22 credits at 1K, 2K or 4K.

The Starter pack is 600 credits for $9.90 and credit packs never expire. Turning a remix into video costs more: Veo 3.1 Lite is 41 credits per 720p clip, Seedance 2.0 Mini starts at 6 credits per second at 480p. The [pricing page](https://whiskai.net/pricing) has the full table for every model.

## Which option should you use?

- **You want a free, playful experiment and Google's tool is available where you live:** use Google's tool.
- **You need predictable results for products, merch or a series:** a Whisk-style remixer with a model choice and 2K/4K output is a better fit.
- **You want to animate the result:** use a tool that hands the image to a video model in one click.

## Privacy and rights

Only upload images you have the right to use, and ask before remixing photos of other people. On WhiskAI, uploads are used to run your generation. Credit packs and paid plans include a private mode, so results don't appear in the public [gallery](https://whiskai.net/gallery). See the privacy policy for details.

## Try the remix

Open the [WhiskAI image studio](https://whiskai.net/image), drop a subject, a scene and a style, and generate. If you don't have images ready, the gallery's "Use this prompt" button fills the studio with a working example.

---

Source: https://whiskai.net/blog/google-whisk-alternative · Site index for AI agents: https://whiskai.net/llms.txt · Full content: https://whiskai.net/llms-full.txt
