Japanese AI company ai& has launched ai& video, a service that generates new footage from text descriptions, photographs and existing video clips. According to the company's announcement on October 22, 2026, the system can produce a video in as little as 12 seconds using infrastructure operated in Japan. Users can describe a scene in plain Japanese, animate a photograph or extend an existing clip.
The service is available through a web browser and an API, which lets developers connect video generation to their own applications. Supported models can generate audio alongside video. Pricing follows a pay-as-you-go model and varies by model, without requiring a monthly subscription. Enterprise customers can also train dedicated models using characters, footage and other assets they have rights to use.
The launch illustrates how AI video creation is expanding beyond standalone creative applications. Students, educators and small businesses may find such services useful for visual explanations, project demonstrations and early-stage concept development. Generating footage quickly can also make it easier to test different compositions and camera movements without producing every version manually.
According to ai&, uploaded materials used for standard video generation are not used to train models. The company says it will continue adding support for video models while working to improve generation speed and processing efficiency.