Google has enhanced its Veo 3.1 AI video generation model, enabling users to create vertical videos for social platforms directly from reference images. This update allows creators to produce native 9:16 vertical format videos for platforms like YouTube Shorts, Instagram, and TikTok, eliminating the need for cropping. The feature has been integrated into the YouTube Shorts and YouTube Create app, streamlining the video creation process.
Introduced in October 2025, Veo 3.1 initially focused on improving audio output and editing controls. With this recent update, the AI model now delivers videos with enhanced character expressions and movements based on reference images, even with concise prompts. Google highlights that the update improves consistency across characters, objects, and backgrounds, allowing users to seamlessly blend various elements for a coherent final output.
These new features are accessible through the Gemini app, while professional users can leverage them via Google’s video editor Flow, Gemini API, Vertex AI, and Google Vids. Furthermore, the update includes an enhanced upscaling feature supporting resolutions up to 1080p and 4K on Flow, Gemini API, and Vertex AI within Google Cloud.
Source: TechCrunch