Kuaishou's next-generation Kling model joins generation, multimodal references and editing in one engine. On SeeVid you can create 3–30 second clips at 720p or 1080p from text, a first/last frame, mixed image and video references, or by editing an existing clip — including 21:9 ultrawide.
Available now on SeeVid — text-to-video, image-to-video, reference-to-video and video edit, 3–30 seconds, 720p / 1080p

Kling AI announced Kling 4.0 on September 28, 2026. Earlier Kling versions treated generation, references and editing as separate steps. Kling 4.0 merges them into one creative model, with three stated goals: realism, control and professional delivery. On SeeVid it is live as text-to-video, image-to-video (first and last frame), reference-to-video (up to 10 images and 5 videos) and video edit (one source clip, optional style images). Clips are 3–30 seconds at 720p or 1080p — long enough for a product spot, a short story beat or a social ad without stitching fragments together.
Kling 4.0 doubles the 15-second ceiling of Kling 3.0. A full scene keeps one consistent camera language, lighting and character performance, instead of being cut into short clips that drift apart.
On SeeVid, Kling 4.0 outputs 720p or 1080p, including 21:9 ultrawide on text-to-video and reference-to-video. The official Kling 4.0 line also describes 4K 10-bit HDR delivery for professional grading.
On SeeVid, image-to-video takes a first frame and an optional last frame. You pin where the shot starts and lands; Kling 4.0 fills in the motion between them.
Reference-to-video accepts up to 10 images and 5 videos (30 seconds in total). Video edit takes one source clip plus optional style images, keeps the original aspect ratio and audio, and can follow the source duration with Auto.
Kling 4.0 targets the gap between an impressive demo clip and footage you can actually deliver. On SeeVid that means longer takes, reference-driven consistency, and instruction-based edits on a real clip — not just one-shot generations.

What you can run on SeeVid today, plus the official Kling 4.0 capabilities behind the model.
Describe the scene, camera moves, lighting and dialogue in plain language. Kling 4.0 turns your prompt into a 3–30 second clip at 720p or 1080p, with 16:9, 9:16, 1:1 or 21:9.
Animate a first-frame still, and optionally add a last frame so the shot lands where you want. Ideal for product turns, character poses and storyboard beats.
Combine up to 10 images and 5 videos (3–30 seconds each, 30 seconds in total) in one job. Point at uploads in the prompt so characters, products and motion stay consistent in a new clip.
Upload one source clip and describe the change. Optional style images, duration Auto or 3–30 seconds, 720p or 1080p. Output keeps the source aspect ratio and audio.
Deliver 720p or 1080p on SeeVid, including 21:9 ultrawide on text-to-video and reference-to-video. The official Kling 4.0 announcement also covers 4K 10-bit HDR.
Dialogue, effects and ambience are generated in dual-channel stereo, and mouth movement tracks speech more closely across languages.
Common questions about Kling 4.0 on SeeVid.
Create 3–30 second clips at 720p or 1080p from text, images, mixed references, or an existing video.