On December 3, in San Francisco (USA), Google Corporation announced that Veo, an Artificial Intelligence (AI) model capable of creating short videos from images and descriptive text, will soon be available to customers using Google Cloud's Vertex AI development platform.

Veo was first introduced by Google last April. This tool can create high-quality videos, with resolutions up to 1080p.

These videos can be up to 6 seconds long, with a frame rate of 24 or 30 frames per second. This model can create videos with a variety of topics, from natural landscapes, objects, people to complex motion effects.

In particular, Veo also supports many different visual and cinematic styles, from landscape, time lapse to editing existing footage.

This model also supports editing specific areas in the video, allowing users to change details without affecting the entire scene.

According to Google, Veo not only stops at creating single videos but also has the ability to edit and combine created videos to create more complete final products.

In addition, Veo also demonstrated the ability to understand and handle visual effects (VFX) quite well, as well as grasp basic physical principles.

This is considered a big step forward in applying AI to video content production, opening up great potential for businesses in many industries.

One of the first customers to use Veo is Quora. The popular question-and-answer site will integrate Veo into its Poe chatbot platform.

Meanwhile, another Veo client, Mondelez International - the company that owns the Oreo cookie brand - will also apply this model to create engaging marketing content with agency partners.

Sharing about this development, Mr. Warren Barkley - Senior Director of Product Management at Google Cloud - said that since Veo was announced, the research team has continuously improved and strengthened this model to serve business customers on the Vertex AI platform.

With Veo support, users can create high-resolution 720p videos, with an aspect ratio of 16:9 for landscape videos or 9:16 for portrait videos.

Veo is expected to compete directly with leading video generation models today such as OpenAI's Sora, or Adobe, Runway, Luma, Meta and many other vendors.

This model has been trained from high-quality video datasets, with a careful selection process to ensure safety and security./.