Google has launched a brand new replace to its Veo AI video era mannequin that ought to make it do a greater job of sticking to prompts and changing photographs into movies. Veo 3.1 is on the market to attempt as we speak via Google's Gemini API and is now additionally powering the corporate's Circulation video editor.
Veo 3.1 builds on the brand new capabilities Google launched with launch of Veo 3 at Google I/O 2025. The brand new mannequin provides higher "immediate adherence," in accordance with Google, and will have a neater time creating movies primarily based on the picture "components" you add alongside your written immediate. Veo 3.1 additionally makes it doable to transform photographs to video and generate audio on the similar time, a functionality that wasn't obtainable with Veo 3.
In Circulation, Veo 3.1 helps no less than a brand new function that offers you finer management over the movies you generate. With what Google calls "Body to Video," Circulation enables you to add a primary and final body, after which generates the video in-between. Adobe Firefly, which is powered by Veo 3, provides an analogous function, however Circulation will have the ability to pull it off and create audio on the similar time. These added audio expertise may even apply to the video editor's skill to increase clips and insert objects into current footage, too.
Based mostly on the samples Google's shared, movies generated with Veo 3.1 nonetheless have an uncanny high quality that appears to differ enormously relying on the immediate and topic. Even when it's lacking among the realism of OpenAI's Sora 2, although, the corporate's resolution to attempt to make Veo extra helpful to individuals who truly work with video quite than a supply of social media spam is a welcome transfer.
This text initially appeared on Engadget at https://www.engadget.com/ai/googles-veo-31-is-better-at-generating-videos-from-images-220829129.html?src=rss