Google's Next-Gen AI Video Model, Veo 3.1, Surfaces Online
Details of Veo 3.1, Google's updated AI video generation model, have surfaced on the Higgsfield AI platform and Google's own Vertex AI cloud service. While the tool isn't public yet, the leaks point to several significant upgrades over its predecessor.
Key Leaked Features
- 1080p Resolution & Longer Clips: The model is expected to generate videos in 1080p with a duration exceeding 30 seconds—a major leap from the 8-second limit of the previous version.
- Scene Builder for Consistency: A new 'scene builder' feature will likely analyze existing footage to generate subsequent clips, ensuring consistent characters and scene elements across shots.
- Multi-Prompting Capabilities: Veo 3.1 may support multi-prompting, allowing users to define different instructions for various scenes within a single video timeline.
An official announcement is anticipated at Google's upcoming 'Gemini at Work' event, which will showcase the latest AI solutions for business and productivity.
192
Comments
Please log in to add comments.
Loading comments...