Google has updated Veo 3.1, its AI video generation model, with a new feature that allows users to create native vertical videos using reference images. In a blog post, Google said it is also bringing a bunch of improvements to its ‘Ingredients to Video’ feature and the ability to generate videos in 1080p and 4K using upscaling.
Google says the improved Ingredients to Video feature now lets users generate “dynamic and engaging videos” from reference images, even with short prompts. Veo 3.1 will now have better dialogues and storytelling, and is more visually consistent. Talking of consistency, the tech giant claims that your characters will look the same even if you change settings, making it easy to use the same character across multiple scenes. Users can also reuse objects, backgrounds and textures.
Google has also added a new option to the Ingredients to Video feature that lets users generate vertical videos in a 9:16 aspect ratio. This means users will be able to create content for YouTube Shorts and other platforms without losing out on quality or cropping. Also, videos generated by Veo 3.1 can now be upscaled to 1080p and 4K resolutions. Google says its improved 1080p resolution offers a “sharper, cleaner video”, while 4K offers “rich textures and stunning clarity” for large screens.
Veo 3.1’s improved Ingredients to Video feature will be available in the YouTube Shorts app, the YouTube Create app, and the Gemini app. The enhanced Veo 3.1 Ingredient to Video and native vertical format features are also rolling out to Flow, Vertex AI, Google Vids and Gemini API.
All videos generated using Google Veo 3.1 will have a SynthID digital watermark, allowing users to verify if the content they are watching is AI-generated by simply uploading an image or video to Gemini.
Comments are closed.