During I/O 2025, Google gave us a sneak peek at its upcoming advanced video generative model, Veo 3. Google highlighted that this new AI model is more advanced from VEO 2 and is capable of creating cinematic content, immersive experiences, and production-ready animations with single prompts. Now, the company has just opened the gates for access to Veo 3 to developers. They can get access to this model through a paid preview via the Gemini API and Vertex AI.
“People around the world have already generated tens of millions of high-quality videos with Veo 3. It is our first video model to incorporate high-fidelity video outputs and native audio, first with text-to-video and soon with image-to-video,” writes Google in its official blog post.
What Veo 3 offers
So, what exactly is Veo 3? Well, Google describes it as the company’s first video model capable of generating super realistic videos with built-in, synced audio like sound effects, dialogue, and ambient noise all generated together.
Right now, the model is capable of turning text prompts into video, but soon it will also allow users to give prompts with images. Users can use the Veo 3 for creating animated shorts, film scenes, storyboards, or game cinematics.
Some of the core features that set Veo 3 apart from current video generation AI models include:
Synchronised Sound: Veo 3 can natively generate audio, including dialogue, music, and sound effects, all timed precisely with the visuals.
Cinematic Quality: The model also supports high-definition video output with rich textures, lighting effects, and creative detail.
Realistic Physics: Veo 3 can simulate real-world physics for believable animation, enabling users to create videos with elements like flowing water, shadow casting, and natural character motion.
Veo 3 open for developers
Google reveals that animation studios like Cartwheel and Volley are already using Veo 3 to boost their creative work. Cartwheel is using Veo 3 to convert 2D human motion videos into fully rigged 3D animations. Meanwhile, Volley is generating game cut-scenes for an upcoming RPG titled Wit’s End, using Veo 3 to rapidly iterate story-driven sequences with high fidelity.
Veo 3 pricing
While Veo 3 looks cutting-edge and impressive, using it comes at a cost. Google has currently priced Veo 3 at $0.75 per second of video with synchronised audio output. In Indian terms, that’s roughly Rs 64.50 per second.
Meanwhile, Google has also announced a faster, more affordable version called Veo 3 Fast. Google says this model will arrive soon. This budget-friendly model is for users who need a powerful yet cost-effective AI tool for quick video creation.
How to use Veo 3
To get started, developers can use Google AI Studio, which includes a starter app, SDK templates, and integration with the Gemini API.
To maintain transparency, Google highlights that Veo 3 embeds a digital SynthID watermark in all its outputs to ensure responsible AI usage.
– Ends