A picture is worth a thousand words? From now on, it’s also worth a video. Google has officially announced an impressive addition to its AI platform, Gemini: Turning a single image into a rich video with movement and sound using the new tool, Veo 3. This step places Google in direct competition with innovative tools like Kling and Luma, which allow static images to be transformed into animated visuals or viral video clips online.
The new feature allows any paying subscriber to upload a single image, enter a short description in a text prompt, and quickly receive an eight-second video with sound. The result: Surreal, funny, or dramatic animations, depending on your request.
The Veo 3 AI model that powers the new feature was first unveiled at the recent Google I/O developers conference. This is the third version of Google’s video engine, and for the first time, it now includes audio support — significantly expanding its capabilities in cinematic content creation.
Demonstrations shared during the tool’s launch included, among others, an image of a simple cardboard box that was transformed into a surreal video — at times containing an elevator, at times a floating ship, followed by a rock concert, and even a tiny mouse cooking in a kitchen.
Google reports that over 40 million videos have already been generated using Veo 3, both via the Gemini app and through Flow — the company’s AI-based video creation platform.
In its initial phase, the new service is available only to Google AI Pro subscribers (priced at $19.99 per month) and AI Ultra subscribers (priced at $249.99 per month), including in Israel. The tool is being gradually rolled out, so not all subscribers may see it immediately. Usage requires a personal Google account and a minimum age of 18. Currently, the service is only available through the Gemini website on desktop browsers, but according to Google, a mobile app version is expected soon.
Alongside the excitement, Google is emphasizing caution: every video produced with Veo 3 includes a visible watermark labeled “Veo” in the bottom-right corner, as well as a hidden watermark using SynthID technology to identify the content as AI-generated. The company notes it is continuing to thoroughly examine how to prevent misuse of the technology, while developing content policies and enforcement measures against the creation of false, violent, or problematic materials.