Kling AI is a generative AI service for video and images, run by Kling AI Pte. Ltd. Its models turn a text prompt, a still image or several reference images into a short video clip. The newest video model, Kling 4.0, generates sound together with the picture.
The video functions include text-to-video, image-to-video (which animates a still picture), multi-image-to-video, video extension, avatars, lip sync, video effects, and audio made from text or from a video. Separate image models, such as Kling Image 3.0, make pictures from text or from other images at up to native 4K resolution, and a virtual try-on function dresses a person in a photo in other clothes. People use the models in a web studio, where a Canvas workspace with an AI agent writes scripts and storyboards, or from an AI agent through MCP and a command-line tool for batch work. Developers call the same models through the Kling API.
The web studio is paid with membership plans and credits. The API is pay per use: each second of generated video costs a number of units, and the price depends on the model and the resolution.