- Google’s video generation model got a big upgrade
- Announced on Google I/O, VEO 3 can combine audio and video in its output
- This is an ultra and only American feature for now
AI video generation tools such as Surah and Pika can create dangerous realistic bits of video, and with considerable effort, you can collect these clips to make a short film. One thing they can’t do, though, produce audio simultaneously. Google’s new VEO 3 model can, and it can be a game changer.
Announced on Tuesday in Google I/O 2025, VEO 3 The third generation of the powerful Gemini video generation model. With the correct indicator, it can produce videos that include sound effects, background noise, and, yes, dialogue.
Google briefly showed this ability for the video model. The clip was a CGI grade animation of some animals in the jungle. Sounds and videos were in perfect sync.
If the demo can be converted to real -world use, it represents a remarkable tiping point in the production space of AI material.
“We are emerging from the silent period of video generation,” said Demis CEO, CEO of Google Deep Mind.
Lights, camera, audio
That’s not wrong. So far, no other AI video generation model can supply audio with video output, or any kind of audio.
It is not yet clear if View 3, which should be able to outpit 4K video like your predecessor, View 2, the current video generation leader in the video Quality Department should be able to overtake the Openi Surah. Google, in the past, has claimed that the VEO 2 is an expert in creating a realistic and permanent movement.
Regardless of, output that fully developed video clips (video And Audio) can immediately make VEO a more attractive platform.
It’s not just that the VEO can handle 3 dialogue. In the world of film and TV, background noise and sound effects are often the job of folly artists. Now, just imagine that if you just need to describe the sounds you want behind and are connected to the action, and it all does it out, including video and dialogue. This is a work in which dynamic people take weeks or months.
In a release on the new model, Google recommends that you “tell a short story in your gesture, and the model gives you a clip back that brings it to life.”
If VEO 3 indicators and output minutes or, ultimately, can follow the permanent video and audio of hours, it will not be too late when we are watching the first dynamic feature fully developed by VEO.
View is straight today and is also available in the United States as part of the new ultra -tire (9 249.99) in the Gemini app and as part of the new flow device.
Google also announced some updates in its VEO 2 video generation models, including the reference to the reflection, camera controls, portraits you provide to landscaping, and the ability to erase Objects Aid and Erase.
Techradar
♬ Original sound – tech Radar