Besides recently launching the new video generation model Veo 2, Google has also introduced a new generative AI image generation experimental project called “Whisk.” Unlike the image generation tools currently on the market, Whisk can generate high-quality images without lengthy, detailed text prompts. You can upload a subject image, a scene image, and a style image, and Whisk can use these to create unique works. However, the explanation might be abstract, so today we’ve prepared images made by internet experts using Whisk to show everyone what Google’s new image generation experimental project Whisk can actually do.
Mix and match themes, scenes, and styles however you like—Google’s latest AI image generation service, Whisk, is a blast to use.
At the same time as releasing the new video generation model Veo 2, Google also launched a new generative AI image generation experimental project called “Whisk.” Google says Whisk can generate the images you want without detailed prompts. Simply upload three images containing the subject, scene, and style, and Whisk can generate content that matches your requested style.
Currently, Google has only opened up Whisk to US users, but in practice it works if you use a VPN to connect to the US. Here are the image results generated by using Whisk from another region—just note that prompts need to be written in English. If you’re interested in trying it yourself, feel free to bypass the region restriction. And if you just want to see what it can do, you can check out works created by other pros using Whisk.
Go to the official Whisk page (requires cross-region access)
As mentioned earlier, only three images are needed to generate content that achieves the desired effect, for example, as shown in the image below:
People might be thinking it’s a bit boring if it’s just like this, but if that’s the case, you can learn from netizens. @ring_hyacinth The method is to first use Whisk to create images that match your ideas, then use 可靈 AI to convert the images into videos, and finally add sound effects to the video.
忽然发现可以用 Google 最新的 Whisk 做出这种巨物玩法！！ pic.twitter.com/CyPaZM3xCY
— Ring Hyacinth (@ring_hyacinth) December 17, 2024
In this way, you can get cute videos like Snorlax sprinkling snow, Pikachu charging the Skytree, Psyduck sleeping in the Roman Colosseum, and so on:
持续制作这个系列！
卡比兽：你再不来，我要下雪了！ pic.twitter.com/qoFBlmTSDJ
— Ring Hyacinth (@ring_hyacinth) December 21, 2024
皮卡丘给晴空塔充电！
太好了是皮卡丘！ pic.twitter.com/snioBXr8bs
— Ring Hyacinth (@ring_hyacinth) December 20, 2024
持续制作这个系列！
可达鸭睡在罗马斗兽场！
创意制作：@simonxxoo & @ring_hyacinth pic.twitter.com/dcxhE3SdsS
— Ring Hyacinth (@ring_hyacinth) December 23, 2024
That concludes the introduction and usage examples of Whisk, Google’s new generative AI image generation experimental project. If you’re interested, you can use a VPN to access Whisk’s service from another region. Or watch the video below to learn more:
Source: KOCPC Chinese