Google announced on Tuesday an update to its Veo 3.1 AI video-generation model, introducing native support for vertical video formats and enhancing its capabilities for creating content from reference images. The changes are aimed at users creating content for social media platforms and are being rolled out across several Google products.
The new features build upon the initial release of Veo 3.1 in October 2025, which provided users with improved audio output and more detailed editing controls over previous versions.
Vertical Video and YouTube Integration
To better serve creators on platforms like YouTube Shorts, Instagram, and TikTok, Veo users can now generate videos directly in the 9:16 aspect ratio. This functionality eliminates the need for manual cropping of standard horizontal videos. Google is also integrating the feature directly into the YouTube Shorts creation tools and the standalone YouTube Create application.
Enhanced Generation from Images
The update significantly improves how Veo 3.1 processes reference images to generate video. The model now produces more dynamic character expressions and movements, even from shorter text prompts. Google stated the changes also lead to better consistency for characters, objects, and backgrounds throughout the generated video. Additionally, users have the ability to blend different elements like characters and backgrounds to produce a cohesive final output.
Upscaling and Platform Availability
A new video upscaling feature has also been introduced, allowing for output in 1080p and 4K resolutions. This particular tool is available through professional-grade platforms, including Google’s video editor Flow, the Gemini API, and Vertex AI. General access to the updated Veo 3.1 features is available in the Gemini app, while professional users can access the full toolset through Flow, the Gemini API, Vertex AI, and Google Vids.