Google is pushing for vertical AI videos with Veo 3.1’s new feature

Key Highlights: 

• Google has just launched Veo 3.1, an update that adds the “Ingredients to Video” feature, which can now generate vertical AI videos using portrait images. 
• The feature supports 9:16 aspect ration, eliminating the hassle of cropping when uploading the videos to social media platforms. 
• The update also improves identity consistency, storytelling, and scene flow. Moreover, upscaling has been bumped up to 1080p and 4K, alongside AI content transparency. 

Vertical videos, which first launched with Snapchat, came into the mainstream after TikTok introduced the For You feed. Now, most social media platforms have adopted this format. On Instagram, you have Reels. On YouTube, you have Shorts, and so on and so forth. Most platforms now drive traffic from short-form content, or rather, short-form video content, I would like to say. That’s also why you see more and more AI-generated videos on your feed these days.

Google Veo 3.1’s new “Ingredients to Video” feature generates vertical videos using portrait images 

Google is quite aware of this. This is why the company has updated its AI video generation tool, Veo, with the latest update. With Veo 3.1, Google has introduced a new feature called “Ingredients to Video,” which uses portrait images to create vertical AI videos that fit right into social media platforms’ existing canvases for vertical videos.

The feature appears to help creators push more and more AI videos to drive traffic and revenue from platforms like TikTok, Instagram, Facebook, and more. Google, in its The Keyword blog post, added that the feature now supports a 9:16 aspect ratio for videos. All you are required to do is throw in some images, or rather portrait images, into Veo, and it will generate vertical videos that are expressive and look more natural, without compromising on details and composition.

For the first time ever, Veo can generate full-screen portrait videos by default. One of the biggest pain points of AI use has been the consistency of elements within videos. Google is taking care of identity consistency with the Veo 3.1 update. The company notes that generated characters and elements within vertical videos are more recognizable across scenes. That applies even when locations change in the video.

1080p, 4K upscaling and AI content transparency with SynthID

That’s not all; the videos generated using Veo 3.1’s Ingredients to Video feature retain backgrounds, objects, and textures across the clip. Aside from all the changes discussed, Veo 3.1 also brings upscaling at up to 1080p and 4K. The improved 1080p output is catered to editing workflows. 

Meanwhile, the 4K upscaling is meant to help professionals who expect top-notch clarity and texture within the videos. Google says that Veo 3.1 access is available in the Gemini app, YouTube Shorts, and YouTube Create for regular users, whereas professional and enterprise users can find it inside Google Vids, the Gemini API, and Vertex AI.

Apart from all this, Google is also making sure that transparency remains one of the most important aspects in the AI era. As a result, videos generated using Veo will now carry SynthID watermarks. In short, you can verify whether a video was created using Google AI tools, like Gemini.