Gemini now lets users generate videos from a single image.
On Thursday, Google announced image-to-video support with Veo 3 video generator on the Gemini app. As of today, you can upload an image to Gemini and turn it into a video with audio, using simple text prompts. Image-to-video tools like Luma and Kling have become popular ways of animating memes or making a still image come to life. Now, Google is doing that too with Gemini.
Powering this tool is Veo 3, which launched at Google I/O this past May. Unlike earlier versions, Veo 3 has audio support, which the internet quickly picked up on. AI-generated videos of news broadcasts and on-the-street interviews went viral for their realism — and misinformation concerns. On that note, Google said all videos generated by Veo 3 include a visible watermark and an invisible SynthID watermark. Users with an AI Pro or AI Ultra subscription will notice the "Videos" option in the tool menu, which they can select, then upload an image and add a description in the prompt box.
In the demo video shared with the announcement, an image of a cardboard box transforms into a video of the box containing, at various times, an elevator, a model ship floating on water, a rock concert, and a mouse in a tiny kitchen. Google said over 40 million videos have been generated through Veo 3 on the Gemini app and Flow, its AI filmmaking platform.
To use this tool, you have have a paid subscription to Google AI Pro which costs $20 a month or Google AI Ultra which costs $250 a month. Users must also be over 18 years old and have a personal Google account, since it doesn't work with company or school accounts.
