Google finally brings the Veo 3 model to the Czech Republic. You can create videos with it directly within Gemini

Google is officially launching its most advanced video generation model, Veo 3, for Czech users with a Google AI Pro subscription. The technology, which was first introduced during this year’s Google I/O developer conference as part of a major update to the Gemini AI service, brings a revolution in AI-generated content. The new feature stands out primarily for its ability to create realistic 8-second videos with integrated sound directly from text descriptions.

Better realism with native sound

The most significant innovation of Veo 3 is native sound generation, which allows adding sound effects, ambient noise, or even dialogue to videos – all directly from text input. The model also brings improved motion physics and much more realistic visual processing. Compared to its predecessors, Veo 3 boasts significantly better adherence to instructions from the prompt, thus more accurately following the desired scenes and actions.

The video generator can process even very detailed prompts with precise descriptions of scenes, characters, and desired actions. Users can define specific shots, camera movements, or even specific dialogues that characters should speak in the video.

After entering a prompt in Gemini, the service will create an 8-second video in 720p quality, which you will receive in MP4 format with a 16:9 aspect ratio. You can get a maximum of three such outputs per day.

Unprecedented creative control

Veo 3 offers advanced options for precise control of generated content. Users can use reference images for more accurate visualization of scenes or characters, define precise camera and object movements, or specify the first and last frame to create smooth transitions.

Other interesting features include outpainting (extending the shot beyond the original frame), adding or removing objects from generated videos, and controlling character movement using one’s own body or voice. Veo 3 also allows maintaining character consistency across different scenes thanks to reference images.

Safety and availability

Google emphasizes the responsible use of generative technologies, which is why all videos created using Veo 3 are marked with a SynthID watermark, allowing identification of AI-generated content. The company has also implemented various safety measures, including blocking harmful requests and results.

In the Czech Republic, Veo 3 is available as part of the Google AI Pro subscription. Users can try the new model through the Gemini web or mobile application.

What creative projects would you like to create using Veo 3?

Source: press release, Google (1, 2)