At Tech Valdemor, OpenAI’s Sora 2 has been dominating headlines lately, our comparison revealed that Google Veo 3 remains the more capable video-generation model overall. Now, Google appears to be extending that lead with the official launch of Veo 3.1, unveiled on Wednesday.

In addition to the standard Veo 3.1 model, Google has introduced Veo 3.1 Fast, a lighter and more efficient version optimized for quicker performance. Both versions are now accessible across Gemini, Vertex AI, and Google Flow—Google’s AI-driven video editing platform designed specifically for creating and refining AI-generated videos.

So, what’s new in Veo 3.1? As its incremental naming suggests, this isn’t a complete overhaul of Veo 3, which debuted only a few months ago. However, Google has made several meaningful improvements. The company claims Veo 3.1 delivers richer audio, stronger narrative understanding, and enhanced realism, particularly when it comes to replicating authentic textures and visual details. These updates aim to make generated videos more cohesive, lifelike, and contextually accurate.

Some existing Veo features have also received notable upgrades—especially through the addition of integrated audio capabilities. One such feature is Ingredients to Video, which allows users to upload reference images that Veo uses to generate tailored scenes. For instance, by uploading a photo of a character and a background, Veo can create a video featuring that character within the chosen setting. This feature, now available with audio, is accessible through Gemini, Vertex AI, and Google Flow.

Another upgraded tool is Scene Extension, which enables users to seamlessly expand an existing video clip. This feature now supports audio within the Gemini API (though not yet in the Gemini app) and is also available in Flow.

Furthermore, Google is enhancing Flow with new precision editing tools, allowing users to insert or remove objects within scenes while maintaining natural and realistic results. Although these tools are still being evaluated, they promise a higher degree of creative control and visual fidelity.

In summary, Veo 3.1 is now live across Google Flow, Gemini, and Vertex AI. However, access to some features depends on their platform rollout stage, and a Google AI Pro subscription is required to use the latest models.