
Create 3–30 second videos with Kling 4.0, with text-to-video, first-and-last-frame image-to-video, and reference-to-video from images and videos at up to 1080p.
See published videos made with Kling models, borrow ideas for motion, camera work, and pacing, then create your own take with Kling 4.0.
Turn reference images into cinematic motion for products, social clips, storyboards, and visual concepts.

ultra realistic high fashion photograph of Emilia Clarke sitting on fresh white snow outdoors in winter. Recognizable facial structure and likeness of Emilia Clarke: oval face shape, soft youthful features, expressive almond-shaped green eyes, naturally full lips, straight petite nose, balanced facial symmetry. Natural winter blush on cheeks and nose. Calm confident expression with a subtle closed-mouth smile, looking directly into the camera. She is seated on the snow with legs bent forward, slightly apart, hands placed behind her for support. Camera angle slightly above eye level, gently looking down. She wears a fitted ribbed sleeveless zip-up yellow bodysuit and pink cable-knit thigh-high socks. Elegant minimal winter fashion styling. Slim, petite, feminine body proportions. Fair-to-light skin tone with natural texture. Background is a snow-covered ground with a dragon partially visible. Natural overcast winter daylight, soft diffused lighting, no harsh shadows, cinematic realism, high skin detail, realistic color grading.

Lens segments combine and move back. Left and right body sections close together. Upper body section moves down. Annotations smoothly appear.

A premium dessert product photograph of a luxury chocolate bar partially unwrapped and centered against a rich warm brown seamless studio background. The chocolate has glossy texture, crisp edges, and a high-end confectionery appearance. Soft cinematic studio lighting, subtle shadows, ultra-sharp focus, premium food advertisement style, hyper realistic, 8K.

isometric office, people working and talking, slow rotation around the isometric scene
Choose duration, frames, references, resolution, and aspect ratio to fit each shot.
Pick any length from 3 to 30 seconds, so one generation can hold a long single take instead of several stitched clips.
Describe the scene, subject action, and camera movement in text to generate a paced video clip.
Upload a first frame, and optionally a last frame, to control where the shot starts and where it lands.
Reference to video accepts up to 10 images and 5 videos, so you can carry a character, product, or motion style into a new shot.
Test ideas at 720p, then switch to 1080p for the final cut. Credits scale with resolution and duration.
Text to video and reference to video support 16:9, 9:16, 1:1, and 21:9. Image to video follows the uploaded frames.
Common uses include long single takes, character-led scenes, and short-form openers.
Describe the opening, the main action, and the ending in order, and let one clip carry the whole moment without cuts.
Up to 30 seconds
Upload character and setting images as references to keep the same subject and look while you try different actions.
Reference images
Generate a vertical 9:16 hook for social posts, then extend the best direction into a longer version.
9:16 vertical
Describe the opening, the main action, and the ending in order, and let one clip carry the whole moment without cuts.
Up to 30 seconds
Upload character and setting images as references to keep the same subject and look while you try different actions.
Reference images
Generate a vertical 9:16 hook for social posts, then extend the best direction into a longer version.
9:16 vertical
Describe the opening, the main action, and the ending in order, and let one clip carry the whole moment without cuts.
Up to 30 seconds
Generate a vertical 9:16 hook for social posts, then extend the best direction into a longer version.
9:16 vertical
Describe the opening, the main action, and the ending in order, and let one clip carry the whole moment without cuts.
Up to 30 seconds
Generate a vertical 9:16 hook for social posts, then extend the best direction into a longer version.
9:16 vertical
Upload character and setting images as references to keep the same subject and look while you try different actions.
Reference images
Upload character and setting images as references to keep the same subject and look while you try different actions.
Reference images
Describe the opening, the main action, and the ending in order, and let one clip carry the whole moment without cuts.
Up to 30 seconds
Describe the opening, the main action, and the ending in order, and let one clip carry the whole moment without cuts.
Up to 30 seconds
Upload character and setting images as references to keep the same subject and look while you try different actions.
Reference images
Upload character and setting images as references to keep the same subject and look while you try different actions.
Reference images
Generate a vertical 9:16 hook for social posts, then extend the best direction into a longer version.
9:16 vertical
Generate a vertical 9:16 hook for social posts, then extend the best direction into a longer version.
9:16 vertical
Answers about duration, references, resolution, and credits.
Need a different answer? Contact support@geminivideos.com.
Any length from 3 to 30 seconds, chosen by the second.
Text to video, image to video with a first frame and an optional last frame, and reference to video with up to 10 images and 5 videos. Reference videos need to be at least 3 seconds each and 30 seconds in total.
Choose 720p or 1080p. Text to video and reference to video support 16:9, 9:16, 1:1, and 21:9; image to video follows the uploaded frames.
Credits depend on duration and resolution. When you add reference videos, a lower per-second rate applies, and the billed seconds include the output plus the longer of the reference videos or the output. The Generate button shows the exact credits before you start, and credits for failed generations are returned automatically.
On AI Video Generator, Kling 4.0 supports up to 30 seconds (Kling 3.0 supports 15), reference to video with images and videos, and the 21:9 aspect ratio. Kling 3.0 still offers motion control and a sound on/off option. You can switch between them in the same generator.
Subscribe to unlock all video and image models, plus more services.
Billed $169 USD / year
Includes
Billed $339 USD / year
Includes
Billed $856 USD / year
Includes
Validate the idea with a short 720p clip, then generate a longer, sharper full version.