Kling AI Video Generator | Text & Image to Video
Stop guessing which model to use. Pick the right Kling model for the job — Kling 3.0, O3, Turbo, Motion Control and native 4K — and generate from one place, with the credit cost shown before you generate.
AI Video Models
Kling 3.0
HOTCinematic video from text / image. standard (720p) · pro (1080p), audio support, multi-shot. 3–15s.
Kling 3.0 4K
4KNative 4K version of Kling 3.0. Cinematic video up to 4K from text / image. Audio & multi-shot. 3–15s.
Kling O3
NEWVideo from text, image & reference image (up to 4). standard · pro, audio support. 3–15s.
Kling O3 4K
4KNative 4K version of Kling O3. Up to 4K video from text, image & reference image. Audio support. 3–15s.
Kling 3.0 Turbo
Fast video generation. standard 720p · pro 1080p, text / image input. 3–15s (5s default).
Kling 3.0 Motion Control
Motion control that transfers a reference video’s motion onto a single image. 720p/1080p, up to 10s (image) / 30s (video).
Kling 2.6 Motion Control
Motion control that applies a reference video’s motion to a character image. 720p/1080p, up to 10s / 30s.
Video Upscaler
common.model_catalog.upscaler
What you can do
Text to video
Write the subject, motion, camera and mood into one prompt and get a cinematic clip.
Image to video
Start from a reference frame (start/end frames supported) and direct the motion.
Native 4K & audio
Kling 3.0 and O3 generate up to native 4K with synced audio in the same pass.
Motion control
Transfer a reference video's motion onto a character image with the Motion Control models.
How it works
Pick a model
Kling 3.0 for finals, O3 or Turbo when cost matters, Motion Control to transfer movement.
Write the shot
Describe the subject, action, camera, lighting and style in one clear scene.
Set resolution & length
Choose 720p, 1080p or 4K and a length from 3 to 15 seconds.
Check credits & generate
Confirm the credit estimate, generate, and download the result.
Video generator FAQ
Can Kling AI generate 4K video?
Yes. Kling 3.0 and Kling O3 offer a 4K quality option alongside 720p and 1080p, and there are dedicated native 4K models (Kling 3.0 4K and Kling O3 4K).
What input types does the video generator support?
Text-to-video, image-to-video (start/end frames), reference-to-video on Kling O3, and motion control (apply a reference video's motion to an image).
Does Kling support audio?
Yes. Kling 3.0 and O3 generate native audio when you enable the audio option. Audio is required when you use multi-shot.
How long can a video be?
From 3 to 15 seconds. In multi-shot mode each shot can run 1 to 12 seconds within the same total.
How is the cost calculated?
Per second of generated video, based on the model and resolution. The generator always shows a credit estimate before you generate.
Start creating with Kling
Pick a model, write your shot, check the credits and generate.
Open the generator