Google’s Gemini AI can now turn your prompts into cinematic videos, here’s how

Google is giving its Gemini Advanced users a powerful new tool: video generation. Starting today, subscribers to the Google One AI Premium plan can access Veo 2, a cutting-edge video model that turns plain text—or even an image—into an animated, high-res video clip in seconds.

Veo 2 is built to impress. Google says it understands real-world physics and human motion better than ever before, delivering smoother animations, richer detail, and more lifelike scenes. The results? Eight-second, 720p MP4 videos in landscape format that feel surprisingly polished for something you generated with a sentence.

You can access Veo 2 directly from the model dropdown in Gemini on both web and mobile. Describe what you want—whether it’s a cinematic slow-motion dragon chase or a quirky animated short about dancing robots—and Veo 2 brings it to life. As with all generative tools, the more vivid your prompt, the better the output. And yes, there’s a monthly cap on how many videos you can create, but Gemini will give you a heads-up as you near it. ALso, you need to be a Gemini Advanced user — meaning you must be a paid subscriber of Google One AI Premium plan.

Google is also integrating Veo 2 into Whisk, the Google Labs experiment that previously let users generate and remix images. Now dubbed Whisk Animate, the platform adds video capabilities, letting you take your image prompts and animate them using Veo 2’s engine. It’s a creative playground for those who like to prototype fast and visual, says Google.

Sharing is frictionless too. With a single tap, you can push your creations to TikTok or YouTube Shorts, making Gemini a powerful tool for content creators looking to stand out with AI-driven visuals.

The rollout of video generation begins today and will continue over the coming weeks for Gemini Advanced users globally.