Google has significantly enhanced its Veo 3 AI model by integrating image-to-video generation capabilities directly into the Gemini app. This new feature allows users to transform static images into dynamic video clips, complete with sound, marking a notable advancement in AI-powered content creation and expanding the creative possibilities for Gemini subscribers.
Veo 3's New Image-to-Video Power
Google announced that its Veo 3 AI video generator, accessible through the Gemini app, now supports image-to-video generation. This functionality was initially introduced in Google's AI-powered video tool, Flow, launched at the Google I/O developer conference in May. The expansion to Gemini makes this powerful feature more widely available to eligible users.
Accessibility and Usage
The image-to-video generation feature is currently available to Google AI Ultra and Google AI Pro subscribers. Users can generate up to three videos per day. To create a video:
Open the Gemini app.
Select the "Videos" option from the tool menu in the prompt box.
Upload a photo.
Describe the desired audio and scene in the prompt.
Within minutes, the tool generates an 8-second MP4 video in 720p HD, complete with motion and sound. These videos can then be downloaded or shared.
Rapid Adoption and Safety Measures
Since its release seven weeks ago, users have generated over 40 million videos across the Gemini app and Flow tool, demonstrating the feature's popularity and utility. Google has also implemented robust safety measures for all AI-generated content:
Visible Watermark: Every generated video includes a visible "Veo" watermark.
Invisible SynthID Watermark: An invisible SynthID digital watermark is embedded, developed by Google DeepMind, to identify AI-powered digital artefacts. Google previously released a tool to detect content containing SynthID.
Google emphasises its commitment to safety through rigorous testing, including "red teaming" to proactively identify and fix potential issues. Users can also provide feedback directly within the app to help improve the experience and safety over time.
Key Takeaways
Google's Veo 3 AI model now supports image-to-video generation within the Gemini app.
The feature is available to Google AI Ultra and Google AI Pro subscribers, with a limit of three creations per day.
Users can transform static images into 8-second, 720p HD videos with accompanying sound.
Over 40 million videos have been generated using Veo 3 across Gemini and Flow in the past seven weeks.
All generated videos include visible "Veo" and invisible SynthID watermarks for transparency and identification of AI-generated content.
Sources
Google adds image-to-video generation capability to Veo 3, Yahoo Finance
