IT Home On October 16, Kuaishou Kling AI announced last night that Kling 4.0 is officially fully available on klingai.com.
4K, 1080p 10-bit HDR format output is about to launch
The creation page synchronization has been fully upgraded, offering a more flexible and free-flowing creation experience.
Kling 4.0 ability access MCP / CLI, reaching more creative scenarios
API services for enterprise customers and developers will be launched soon
According to the introduction, Kling 4.0 has seen significant improvements in visual realism, creative control, and narrative completeness. It has made breakthroughs in multiple aspects such as audio-visual performance, multiple key frames, mouth shape matching, text generation, and creative replication.
The main upgrade details for IT Home are as follows:
A comprehensive upgrade has been made to high-speed movement, continuous actions, tracking, and orbiting effects. The stability and continuity of the camera lenses have reached more professional standards.
Supports high-quality audio and dual-channel stereo, providing greater depth and layering of sound, closer to the realism of real-time audio, creating a completely new “immersive” auditory experience
Further refined the accuracy of the character mouth shape matching, making the coordination between voice, dialogue, and performance more natural, and giving the character portrayal a greater “human-like” feel
Supports 4K and 1080p 10-bit HDR format output, with richer color levels and smoother light and dark transitions; also supports a 21:9 ultra-wide aspect ratio.
Up to 10 images, 5 videos, and 7 subjects can be entered at a time for use in the same generation task combination. Different reference materials work together, resulting in a more accurate portrayal of characters, actions, shots, and sounds, which reduces the cost of repeated debugging.
Supports various inputs such as text, images, and subjects. It allows adding, modifying, and deleting the subjects and background in the original video. You can also adjust the video’s style, weather, colors, materials, shot angles, and other aspects.
It supports generating videos up to 30 seconds at a time, and supports long shots and continuous storytelling, allowing emotions, actions, and shots to naturally develop along the same narrative line.
Supports up to 10 multi-keyframe inputs, allowing free setting of key frame nodes, and precise control over character states, scene changes, and plot nodes.
Support input length up to 8000 tokens, and reasonably supplement, expand, and refine the prompt words, introducing creative judgment like that of a director.
Supports visual styles such as cinematic, animation, fantasy, pixel art, and felt.
Supports multiple languages such as Chinese, English, Japanese, Korean, Spanish, Portuguese, German, French, and Hindi, as well as various accents and dialects like Beijing, Taiwan, Northeast China, Sichuan, Cantonese, American English, British English, and Indian English. It also supports generating multi-language text, Emoji, and Logos.
Supports multiple continuous video recordings, with a maximum length of 2 minutes.
Related Reading: