Google unveils the latest version of its video generation model Veo- Veo 2 in its latest blog post that can achieve state-of-the-art results just from a text prompt. It is capable of generating fully professional 4K cinematic outputs lasting up to a minute with the most intricate details and customization. It is currently available in VideoFX, ImageFX, and the latest Google Lab experiment, Whisk. 

The video generation tool aims to serve people ranging from developers to filmmakers. Google claims that the model has achieved state-of-the-art results against leading models upon comparison across the subjects and styles. 

Understanding of the model 

As per the blog by Google, the latest version of Veo understands the basics of cinematography like genres, lenses, cinematic effects, etc. Apart from these, its understanding of real-world physics and the nuances of human movement and expression are an added advantage for a Gen-AI model. 

Read about Google Whisk here: Is Google Whisk available in India? A transformative Gen-AI tool that generates images through image prompts

Quality of video output 

The model produces 4K cinematic videos up to a minute. The model produces fewer hallucinations and unwanted details like- fingers and objects in the video, making the output look more realistic and higher in quality. 

At this point, it is highly recommended to visit Google's blog and witness the outstanding output delivered by Veo-2 or you can also see it in the Tweet pinned below. 

Today, we’re announcing Veo 2: our state-of-the-art video generation model which produces realistic, high-quality clips from text or image prompts. ????

We’re also releasing an improved version of our text-to-image model, Imagen 3 - available to use in ImageFX through… pic.twitter.com/h6ejHaMUM4

— Google DeepMind (@GoogleDeepMind) December 16, 2024

Invisible watermark 

Similar to other image and video generation models, Veo 2 includes an invisible SynthID watermark in its outputs, helping to identify them as AI-generated and reducing the risk of misinformation and misattribution. It's been included in the model as a layer of safety to prevent any unforeseen or unwelcoming instances in the future. 

Interested in AI? Read about Deep Research, Google's latest AI innovation, at: Google unveils Deep Research in Gemini AI to automate research and generate reports in seconds

Availability 

The all-new video generation model, Veo 2, is now available on Google Labs' VideoFX tool, where users can sign up to experience its magic. Additionally, it is set to expand to YouTube Shorts and other products by next year.