What we know
The update changes how videos generated from images behave: facial expressions, character movements, and overall scene coherence have been noticeably improved. Backgrounds, objects, and textures are now rendered more consistently, and the model reproduces scenes more accurately even from short text prompts.
Google has also integrated the new capabilities directly into YouTube Shorts and YouTube Create. Veo 3.1 was first introduced in October 2025 with enhanced audio and expanded editing tools. The latest update adds improved image-based generation and native support for vertical video.
The features are available through the Gemini app, and for professional users in Flow, via the Gemini API, Vertex AI, and Google Vids. Google Cloud has also updated its video upscaling tools, enabling output at 1080p and 4K resolution.
From an industry perspective, Veo 3.1’s shift toward native vertical video is less about format support and more about acknowledging where real audience attention now lives.
