Kling AI 4.0 is a multimodal video generation model that converts text ideas or static images into short video clips. The tool is designed to produce cinematic scenes, social media clips, or product videos with smooth motion and synchronized sound. Videos can be created with a duration ranging from 3 to 30 seconds, and users can guide the output by providing reference materials such as images, videos, or audio files. The platform supports up to 10 reference images in PNG or JPG format, each up to 30 MB with dimensions between 300 and 3840 pixels per side and a shorter side no larger than 2160 pixels. Up to 5 reference videos in MP4 or MOV format can be used, each between 3 and 30 seconds long, with a combined total of 30 seconds, a maximum file size of 300 MB per video, and a minimum shortest side of 700 pixels. Optional reference audio in MP3 or WAV format can be added, with a duration of 5 to 30 seconds and a maximum size of 50 MB per file; audio requires a 720p output resolution and cannot be combined with the original video sound. Output settings include resolution options, and credits are charged per second of generated video. Typical use cases include creating short narrative scenes, maintaining character or product consistency across shots, and refining existing footage with more natural camera movements or wider framing. The workflow involves adding source media as references, selecting output parameters, and generating the video. Kling AI 4.0 also introduces capabilities for longer scenes, directing key moments, working with richer references, and iterating on previously generated clips. The tool is available through the Kling 4.0 web interface, and users should review the associated terms of service, privacy policy, cookie policy, and refund policy.