- Kling 4.0 is now fully available to all users on klingai.com with the complete feature set.
- Single-shot video generation reaches 30 seconds natively, extendable to two minutes through continuation, enabling continuous narrative storytelling within a single generation pass.
- Output quality targets professional creators with 4K and 1080p 10-bit HDR support, 21:9 ultra-wide aspect ratio, and dual-channel stereo audio with tightened lip-sync accuracy.
- MCP/CLI integration and an upcoming enterprise API indicate Kuaishou's strategy to embed Kling into external production pipelines rather than keep it as a standalone consumer tool.
- The model accepts up to 10 images, five videos and seven subjects in a single task, with an 8,000-token prompt ceiling and auto-refinement features that reduce iterative manual tuning.
Full Public Launch on klingai.com
Kuaishou announced that Kling 4.0 has been fully rolled out on klingai.com, making the complete model accessible to all registered users. The release represents a significant upgrade in visual realism, creative control and narrative completeness compared to previous versions.
The full launch includes the complete feature set spanning visual quality improvements, multi-keyframe control, lip-sync matching, text generation and creative replication capabilities. Kuaishou states that high-speed motion, continuous action, tracking shots and orbital camera movements have been refined to meet more professional production standards.
Longer Generation and Narrative Continuity
Kling 4.0 supports native single-shot video generation of up to 30 seconds, a significant increase over previous versions that typically produced shorter clips. Through multiple continuation passes, creators can extend a single sequence to approximately two minutes, allowing emotions, actions and camera movements to develop along a continuous storyline without stitching separate generations together.
The input side has also been expanded: a single task can now incorporate up to 10 reference images, five video clips and seven distinct subjects working in concert. The prompt limit has been raised to 8,000 tokens, and the system includes an auto-refinement feature that analyzes reference material and suggests compositional adjustments, which Kuaishou describes as introducing 'director thinking' into the generation process.
Professional-Grade Output Specifications
Output resolution now includes 4K and 1080p 10-bit HDR formats with smoother highlight and shadow transitions, alongside support for the 21:9 ultra-wide cinematic aspect ratio. On the audio side, Kling 4.0 generates dual-channel stereo sound with improved depth and layering, and character lip-sync accuracy has been tightened so that dialogue, performance and audio track align more naturally.
Visual style presets span cinematic, animation, fantasy, pixel art and felt aesthetics. Language support covers Mandarin, English, French, German, Japanese, Korean, Spanish, Portuguese and Hindi, along with regional dialects including Beijing, Taiwanese, Northeastern Chinese, Sichuan and Cantonese accents.
Workflow Integration Over Standalone Demo
Beyond generation quality, Kling 4.0 has been integrated with MCP and CLI interfaces, allowing developers and studios to invoke the model programmatically rather than solely through the web interface. An enterprise-facing API is scheduled to open in the near term, though specific pricing and access tiers have not yet been disclosed.
Kuaishou also states that the creation page has been upgraded alongside the model release. The company describes these changes as positioning Kling as a tool embedded in production workflows rather than a standalone novelty, emphasizing a deliberate shift toward 'tool-based' rather than 'toy-based' usage.
Sources & context
Go to the original material. Company claims remain attributed to their sources.
01
