Google Veo 3.1 introduces vertical video and improved reference tools

Google Veo 3.1 specifically targets the growing demand for short-form mobile content by introducing native 9:16 vertical video generation. Previously restricted to widescreen formats, the model now allows users to create high-definition video clips optimized for platforms like TikTok, YouTube Shorts, and Instagram Reels. This update minimizes the need for third-party cropping tools and ensures that the framing and composition of AI-generated scenes are designed specifically for vertical viewing from the initial prompt.

Google Veo reference image and style consistency

Google Veo now features a sophisticated “reference image” tool that significantly improves character and environment consistency. By uploading a single static image, creators can instruct the AI to maintain specific facial features, clothing, or artistic styles across a series of generated videos. This technical advancement addresses a common issue in generative video where subjects often change appearance between clips, making the tool more viable for professional storyboarding and long-form digital storytelling.

Google Veo release and availability details

Google Veo 3.1 is currently rolling out to select creators through the VideoFX platform as part of a phased deployment in early 2026. While Google has not yet announced a standard public “price per minute” for the service, it is integrated into the premium tiers of Google’s AI subscription services. The company plans to expand access to a wider audience of marketing professionals and developers through the Vertex AI platform later this year, providing a competitive alternative to other high-end video generation models.