Kling 1.6

Yes, Kling 1.6 is the major generational release of the Kling AI video model family developed by Kuaishou Technology. It is powered by a 3D Variational Autoencoder and Diffusion Transformer architecture designed for synchronous spatiotemporal compression and physical motion fidelity.

Pro mode outputs native 1080p full HD video with 30 FPS playback, superior facial temporal consistency, and support for complex multi-action prompts across 5 or 10-second durations. Standard mode generates 720p video capped at 5 seconds at roughly half the computational cost, making it ideal for rapid storyboarding and prompt testing.

Use Image-to-Video when you have an existing visual keyframe or character concept and need strict visual continuity, accurate lighting, and preserved facial landmarks. Use Text-to-Video when generating entirely new scenes from scratch without reference imagery, directing the composition through descriptive camera and motion prompts.

For standard generation, crop input images to 960x544 pixels to preserve 1:1 pixel fidelity and avoid automatic resizing or cropping. Ensure your image aspect ratio matches your generation settings within the supported 2:5 to 5:2 range, using clean, high-contrast imagery with well-defined borders.

Kling 1.6 is not designed for rendering crisp legible text or signage, multi-character crowd scenes with overlapping action paths, or static ambient loops, which can induce background warping or frozen subject glitches.