Google’s Veo 3.1 Update Adds Vertical Video Creation Using Reference Images

Google has rolled out an update to its Veo 3.1 AI video-generation model that enables users to create native vertical videos using reference images, a move aimed at improving content creation for social platforms. The update also improves the appearance of expressive, dynamic videos generated from reference imagery.

With this update, creators producing AI-generated videos for platforms such as YouTube Shorts, Instagram, and TikTok can now select a native 9:16 vertical format. This removes the need for cropping and ensures videos are optimised for vertical viewing. Google is also integrating this capability directly into YouTube Shorts and the YouTube Create app.

Google initially launched Veo 3.1 in October 2025, introducing improved audio generation and more detailed editing controls compared with earlier versions of the model.

According to the company, Veo 3.1 now delivers stronger character expressions and more natural movements when reference images are used, even with shorter prompts. The update also improves consistency across characters, objects, and backgrounds. In addition, users can combine multiple characters, environments, objects, and textures to produce a more unified and cohesive video output.

These features are available to users through the Gemini app. Professional users can access them via Google’s Flow video editor, the Gemini API, Vertex AI, and Google Vids.

The update also introduces enhanced upscaling support, enabling videos to be rendered at 1080p and 4 K. This higher-resolution output is available through Flow, the Gemini API, and Vertex AI within Google Cloud.