AI Video Just Got a Lot Closer to the Real Thing: Google Unveils Veo 3.1

Google has rolled out a significant update to its AI-powered video generation services, giving users more advanced and user-friendly tools for creating and editing clips. The key change is the release of the new Veo 3.1 model, which developers say handles tasks significantly better than its predecessors.

Videos created by the upgraded AI model are now higher quality and more detailed, and the AI has a better understanding of what the user wants from their prompt. A special focus was placed on sound, which has also become more refined and realistic.

These improvements are directly reflected in the Flow video editor, which now includes features many users have been asking for. For instance, you can add a new object to an existing clip—like placing a cup on a table or a bird in the sky—and the system will automatically blend it in with the lighting and environment for a natural look. The inverse function—removing unwanted elements from a scene to clean up the composition—will also be available soon.

For those who want more control, other new options will prove useful. You can upload several reference images to define the style and look of your video, or seamlessly merge two different shots with a smooth transition—a task that previously required third-party editing software. And if a generated clip isn't long enough, you can literally "extend" it, with the AI creating a continuation based on the last few seconds, opening up new possibilities for building longer scenes.

Working with video structure has also been simplified. Users can now directly influence the action on screen, making edits at any stage without having to start over from scratch. The updated Veo 3.1 model is already being integrated into various Google services, including the Gemini app and developer platforms. The new features will gradually roll out to all Flow users.