Google is introducing a new feature for paid Gemini AI users that transforms still images into short video clips. Available initially to subscribers of Gemini’s Ultra and Pro plans via the web, the feature is expected to reach the mobile app later this week.
How It Works
Users can create 8-second video clips—with audio—from a single image, enhanced by any text prompt they provide. These videos are exported as MP4 files in 720p resolution and a 16:9 landscape format, according to Bloomberg.
The feature is powered by Veo 3, Google’s latest video generation model unveiled during its I/O 2025 developer conference. Veo 3 is also used in Google’s standalone filmmaking tool, Flow.
Ethical and Technical Safeguards
To prevent misuse, Google restricts the creation of videos featuring publicly identifiable people, including celebrities and political figures. Outputs that promote violence, dangerous activities, or bullying are also blocked.
While there’s no built-in instruction to alter a subject’s appearance, Google acknowledges that face animation and photo-to-video technology is still evolving. At present, the model performs better on objects, nature images, and artwork. Improvements in facial rendering are planned for future updates.
Sundar Pichai Shares Milestone
Google CEO Sundar Pichai highlighted the feature’s debut in a post on X (formerly Twitter):
“Since I/O in May, you’ve created 40M+ videos with Veo 3! Now our new photo to video feature in the @Geminiapp lets you create clips inspired by the world around you,” he wrote, sharing a light-hearted example involving Google’s campus dinosaur mascot, Stan.
Competitive Landscape
By embedding video tools directly in Gemini’s AI chat interface, Google is ramping up its competition with OpenAI and video-focused platforms like Runway. It also faces mounting pressure from Chinese tech firms like Alibaba’s AI venture Manus and Kuaishou, both of which are advancing their generative video capabilities.