Google recently announced that Veo 3, its advanced AI video generation model, is now available for developers through a paid preview in the Gemini API and Google AI Studio. This launch enables developers and creators to produce high-quality videos complete with synchronized dialogue, music, and sound effects, all generated from a single prompt.
Build with Veo 3 in the Gemini API: Google’s Veo 3 now available for developers
With Veo 3, developers gain enhanced artistic oversight, including:
- Synchronized Sound: Veo 3 integrates dialogue, background music, and realistic sound effects in a single pass, eliminating the need for post-production audio work.
- Cinematic Quality: The model generates high-definition video that captures creative details, from intricate textures to subtle lighting effects specified in the prompt.
- Realistic Physics: Veo 3 simulates real-world physics for authentic motion, covering everything from natural character movements to the accurate flow of water and shadow casting.
Early use cases and developer experiments
Since its reveal at Google I/O 2025, Veo 3 has been used to generate tens of millions of videos worldwide, sparking new creative trends and experiments.
Examples of early adopters include:
- Cartwheel developed a system that transforms 2D human videos into fully production-ready 3D animations on rigged characters. Using Veo 3, Cartwheel generates realistic human motion, which is then converted into 3D animations for clients.
- Volley leverages Veo 3 to produce in-game video cut-scenes that advance storytelling. The team uses the model to rapidly iterate and refine sequences for an upcoming RPG titled Wit's End.
Capabilities highlighted
Veo 3 supports a wide range of video generation tasks, from cinematic storytelling to dynamic animations. Key capabilities include:
- Synchronized Sound: Native generation of dialogue, sound effects, and music, aligned with video in one pass.
- Cinematic Quality: Production of HD video capturing detailed artistic elements from provided prompts.
- Realistic Physics: Simulation of authentic motion, such as character movement, flowing water, and environmental interactions.
Example prompts demonstrating Veo 3
- Fluffy Characters in Stop Motion: A colorful kitchen scene made of felt and yarn, where Professor Nibbles, a plump hamster with glasses, nervously stirs a pot that erupts in green slime, causing frantic chaos.
- Mechanical Heart in Desert: An extreme close-up of a turning gear reveals a massive mechanical heart in a desolate desert, with tiny robed figures performing maintenance, emphasizing scale and intricate machinery.
These examples and more can be explored using the SDK template and Starter App available in Google AI Studio. Paid Tier users can remix, copy, and extend samples directly within the platform.
Getting started and pricing
To access Veo 3, developers must select a Google Cloud Project with billing enabled within AI Studio. Pricing is set at $0.75 per second for video and audio output. A Veo 3 Fast option is also planned, promising quicker, cost-effective video generation.
Commitment to responsible AI
All videos created with Veo 3 include a digital SynthID watermark for responsible content generation. Developers are encouraged to explore the official documentation, cookbook, and starter app to familiarize themselves with best practices.
In addition to Gemini API access through Google AI Studio, Veo 3 is also available to subscribers of the Gemini app and Flow, as well as enterprise customers using Vertex AI.