Google has introduced a new photo-to-video generation feature on the Gemini platform, allowing users to animate still images into dynamic, 8-second video clips using the Veo 3 model.
This feature is available to Google AI Pro and Ultra Subscribers today with carefully selected local Ultra Subscribers.
An announcement from David Sharon, Gemini Apps' multimodal generation lead, means integration into existing video tools available in Gemini and Flow, Google's AI film production suite. Users can upload static photos, describe the scene and desired audio, and receive AI-generated videos with motion and sound.
The VEO 3, a third-generation video generation model announced in May 2025, supports this new feature. The model specializes in generating high-fidelity video content from multimodal prompts, and over 40 million video creations have already logged the Gemini app and Flow within seven weeks. Creative applications ranged from surreal rethinking fairy tales to sensible ASMR clips.
Gemini is Google's unified generator AI assistant, available on the web and mobile, and is positioned as Bard's successor. Gemini is integrated with Google's ecosystem of apps and services, targeting both the average user and creative experts.
To access the new features, users can select “Video” from the Gemini prompt interface, upload photos, and provide text describing the desired motion and audio elements. The system generates 8-second clips based on the input. Google suggests use cases such as encouraging personal photography, sketching or landscape experiments, and realizing animation and artwork for nature scenes.
In terms of safety, Google says it has implemented layered safeguards to mitigate misuse. These include red teaming exercises, an aggressive assessment of potential abuse scenarios, and a strict content moderation policy. All videos generated via VEO 3 are marked with both visible AI-generated watermarks and invisible SynthID signatures, which help users and platforms verify reliability. The feedback mechanism is also in place, which helps users evaluate video output and improve the system.
The deployment is limited to subscribers to Google AI Pro and Ultra in a particular country, but the company points out that users who are already experimenting with AI-based storytelling will have access to the same tools.
If you like this article, be sure to follow us X/Twitter And again LinkedIn More exclusive content.
