During its annual developer conference on May 20, Google unveiled Veo 3, its newest video AI model that can also incorporate audio in videos it creates. The company claims that Veo 3 can not only generate realistic videos but also give sound effects, background noises, including dialogues. From seamless transitions to audio generation, here are 10 ways Veo 3 takes AI-driven video generation to the next level.
WE CAN TALK! I spent 2 hours playing with Veo 3 @googledeepmind and it blew my mind now that it can do sound! It can talk, and this is all out of the box… pic.twitter.com/ufplpcZWbq
— Ari K (@arikuschnir) May 20, 2025
The video posted by Ari K shows different characters talking to one another in different accents. From movie characters to cartoons, Veo 3 seems to be accurately generating audio for each character in the video. This is impressive since, just over a year ago, AI video models were struggling to imitate human expressions.
Did someone say 100 men vs a gorilla at a rave dance off? #veo3 pic.twitter.com/CDBmIo0TIG
— Ruben Villegas (@RubenEVillegas) May 20, 2025
The video shows a gorilla dancing in a bar alongside humans. The detail, lighting and movements all look way too realistic.
Video, meet audio. 🎥🤝🔊
With Veo 3, our new state-of-the-art generative video model, you can add soundtracks to clips you make.
Create talking characters, include sound effects, and more while developing videos in a range of cinematic styles. 🧵 pic.twitter.com/5Hfpetfg8b
— Google DeepMind (@GoogleDeepMind) May 20, 2025
Google’s DeepMind official X handle also shared some short clips that show how Veo 3 can deliver dialogues with human-like expressions, simulate physics, and create sci-fi worlds that look incredible.
AI video just made a huge leap with Google Veo 3. Creatives are going to have a field day.
We’ve jumped from Commodore 64 to the first PC on the timeline. pic.twitter.com/oW8WTBWf9c
— Ben Kusin (@bkvenn) May 21, 2025
This video shows different characters saying the famous tongue twister, “how much wood can a woodchuck chuck if a woodchuck could chuck wood?”
#Veo3 further blurs the lines between reality and imagination with audio, stronger text adherence, and richer visual details. pic.twitter.com/Tv5H9Uwzh3
— MBZ (@babaeizadeh) May 20, 2025
Veo 3 also creates stunning and detailed visuals with audio that matches the scene.
You gotta check out this giraffe’s flow! 🎤 #veo3 pic.twitter.com/GUVwDOK5G6
— Emanuele Bugliarello (@ebugliarello) May 20, 2025
As it turns out, Veo 3 can also generate accompanying subtitles in different styles that are also accurate.
10/10 Noodle physics from Veo 3. pic.twitter.com/FP61xlfKzO
— Pizza Later (@Pizza_Later) May 20, 2025
While OpenAI’s Sora struggled with physics at launch, it looks like Veo 3 has no problems imitating how objects realistically interact with one another.
Created with Google Flow.
Visuals, Sound Design, and Voice were prompted using Veo 3 text-to-video.
Welcome to a new era of filmmaking. pic.twitter.com/E3NSA1WsXe
— Dave Clark (@Diesol) May 21, 2025
When Veo 3 is used in combination with Google Flow, the company’s new AI tool for filmmaking, the result is a clip that looks like it’s shot by professional filmmakers.
this is wild
> a man in a music video raps to the camera about generating videos with Veo 3 pic.twitter.com/GZs104nDJY
— fofr (@fofrAI) May 20, 2025
It looks like Veo 3 can not only make realistic-looking videos but also rap.
In a blog post, Eli Collins, Google’s DeepMind product vice president, said, “Veo 3 excels from text and image prompting to real-world physics and accurate lip syncing.” Google’s new AI video tool is currently available for those who are subscribed to Google’s new Gemini AI Ultra plan, which costs $249,99 per month and will soon be available for Vertex AI enterprise platform users.
