Google's Veo 3.1 Is Better At Generating Videos From Images

Google has released a caller update to its Veo AI video procreation exemplary that should make it do a amended occupation of sticking to prompts and converting images into videos. Veo 3.1 is disposable to effort coming done Google's Gemini API and is now besides powering nan company's Flow video editor.

Veo 3.1 builds connected nan caller capabilities Google introduced pinch motorboat of Veo 3 astatine Google I/O 2025. The caller exemplary offers amended "prompt adherence," according to Google, and should person an easier clip creating videos based connected nan image "ingredients" you upload alongside your written prompt. Veo 3.1 besides makes it imaginable to person images to video and make audio astatine nan aforesaid time, a capacity that wasn't disposable pinch Veo 3.

In Flow, Veo 3.1 supports astatine slightest a caller characteristic that gives you finer power complete nan videos you generate. With what Google calls "Frame to Video," Flow lets you upload a first and past frame, and past generates nan video in-between. Adobe Firefly, which is powered by Veo 3, offers a akin feature, but Flow will beryllium capable to propulsion it disconnected and create audio astatine nan aforesaid time. Those added audio skills will besides use to nan video editor's expertise to widen clips and insert objects into existing footage, too.

Based connected nan samples Google's shared, videos generated pinch Veo 3.1 still person an uncanny value that seems to alteration greatly depending connected nan punctual and subject. Even if it's missing immoderate of nan realism of OpenAI's Sora 2, though, nan company's determination to effort and make Veo much useful to group who really activity pinch video alternatively than a root of societal media spam is simply a invited move.