Google launches video-generating model Veo for more customers amid OpenAI's recent Sora leak

Google has recently announced that its video-generating model, Veo, is now available on Vertex AI, Google Cloud’s machine learning (ML) platform for enterprises.

Initially launched in May 2024, Veo can generate high-quality videos from text or image prompts with tools for rapid prototyping and easier workflow. The feature itself is now in private preview and Google said that it’s rolling out “starting next week.”

But it’s not entirely replacing a video production team, though. Google also mentions that Veo is nothing but a “great partner for human creativity” that helps “handle tedious or repetitive aspects of video production.”

“As the first hyperscaler to offer an image-to-video model, we’re helping companies transform their existing creative assets into dynamic visuals,” Google describes the Veo model.

The move come not too long after OpenAI experienced a “leak” (or, perhaps the best way to explain it an “exposure) of its hotly-anticipated Sora model.

A group of artists exposed Sora’s front end in protest of unpaid labor, meaning that non-beta testers who weren’t a part of OpenAI’s selected group could try it. It only lasted for three hours though until the Microsoft-backed company took it down.

The Mountain View tech giant also says that the text-to-image model that powers the Gemini chatbot, Imagen 3, is also coming to the platform. The model had a quiet launch following concerns over historically inaccurate images, so bad that Google had to temporarily pause it.

Imagen 3, which is now available on Google Docs as well, does not allow the creation of “photorealistic, identifiable individuals, depictions of minors or excessively gory, violent or sexual scenes.”