What you need to know
- Google is making its Veo 3 and Veo 3 Fast video generators “generally” available for users in Vertex AI.
- Veo 3 Fast tries to balance speed and Veo’s high quality video generation for users looking for quicker results.
- Google teases that both models will pick up image to video capabilities in another update slated for August in Vertex AI.
There are some huge Veo 3 updates rolling out for interested parties today (July 29), and a major teaser about new—and familiar—capabilities.
Google detailed Veo 3’s updates this morning via a Cloud blog post, informing users that the AI-powered video generator is now fully available in Vertex AI. Outside of this, Google is also introducing a new version, called “Veo 3 Fast.” This Veo model is a little more lightweight for users looking to complete “work that requires rapid iteration and speed.”
Veo 3 Fast is reportedly adept at generating and testing variations of a company’s advertisements “quickly.” This model can also create “video demonstrations” for your products based on the images you’ve provided. Additionally, Google states Veo Fast can create “engaging animated explainers” and even training material for new recruits.
The company speaks about Veo 3 Fast in the same vein as it does Gemini 2.5 Flash. Veo 3 Fast, while being lightweight, has reportedly struck a balance between fast processing times and strong visual quality output.
Google’s Veo series goes further
Google’s Veo 3 Fast model is available in Vertex AI alongside the larger Veo 3 model. The post highlights some ways users can enjoy both AI video generators, such as more storytelling. In Vertex AI, users can set the scene, alongside any audio they’d like to include, all in a single prompt. If you have your characters speaking, Google says what you’d like them to say will be lip-synced, as well as any other background sounds you’d like to include.
On a related note, Veo 3 features native dialogue generation, which helps businesses “connect with an international audience.” While Veo 3 and Veo 3 Fast generate your video once, it will “localize the dialogue for dozens of languages.”
All videos generated arrive in 1080p, giving users a little more confidence in their quality before showing it off.
Regarding that Veo 3 teaser, Google says that both models will receive an image-to-video generation feature in Vertex AI in August. We’ve seen this arrive a couple of times already through an update for Gemini and another for Google Photos, both of which rolled out in July. Now, Veo 3 and Veo 3 Fast will have this feature available for users to experiment with, granting them an eight-second video clip based on the photo you’ve provided.
