Kling Video O1: generate and edit in one model
Kling Video O1 brings up to 18 video skills into one model. Generate videos from text, use images and saved subjects as references, transform existing footage, or continue a scene from a video. You can also combine these capabilities in a single prompt, so you can create a shot and then shape what happens next without switching between separate models.
What is Kling Video O1 AI?
Kling Video O1 is a unified multimodal video model that brings different video tasks into one place. It can generate, transform, reference, and continue video using images, existing clips, saved subjects, and text as inputs. Instead of switching between separate modes for each task, you can combine different instructions in the same prompt and shape the video you want from there.
What Video O1 is built to do
Save a character, product, or prop as an Element and reuse it across shots, while keeping its appearance consistent as the camera and scene change.
Keep the same characters across scenes
Kling Video O1 can use reference images to keep a character, product, or object consistent as you create new shots. Change the camera angle, movement, or scene, and the model works to preserve how your subject looks. You can also reference multiple characters or objects at once, making it easier to create a sequence without their appearance changing from shot to shot.
Generate, edit, and continue your video
Video O1 goes beyond generating a new clip. Describe what you want to change - remove something from the shot, switch the background, restyle the scene, change the weather, or adjust an object - without manually masking or setting keyframes. You can also give it an existing video and generate the shot before or after it, or use the clip’s action or camera movement as a reference for something new.
Kling Video O1 in Picsart's AI Playground
Kling Video O1 is available alongside other video models in Picsart's AI Playground, so you can explore different approaches to the same idea. With Video O1, you can go beyond the first generation and keep shaping the result - edit the scene, change specific details, or build on the video you’ve already created without starting over.
Kling Video O1 is a unified multimodal video model released on 1 December 2025. It covers five categories of skill - image and element reference, transformation, video reference, start and end frames, and text-to-video - adding up to 18 skills, all driven from a single prompt box.
Change video that already exists. Other Kling models generate from a prompt or a reference; Video O1 also removes and adds content, swaps subjects and backgrounds, alters style, color and weather, switches the angle or shot size, and generates the shot before or after a clip you supply.
An Element is a saved subject - a character, a product, a prop - with its own name, description and reference views. Once saved it can be reused across generations, and the model keeps its features consistent from shot to shot however the camera moves.
Between 3 and 10 seconds, and you choose the length. For longer work, Kling 3.0 Omni generates up to two minutes.
Yes. A single instruction can add a subject while changing the background, or restyle a shot while using an element you saved. Kling calls these skill combos.
Yes. Given a reference clip it generates the shot before or after within the same setting, and can reuse the clip's camera movement or transfer its action onto a character you supply.
Yes. Videos generated through Picsart's tools powered by Kling Video O1 can be used for marketing, social media, brand content and other commercial applications, subject to Picsart's terms of use.
More AI models to use
Generate AI visuals with Kling 3.0 Omni
Generate AI visuals with Kling 3.0 Omni
Compare Kling Video O1 with other video and audio models for motion, sound, and campaign work.
Learn how prompts, clips, and model choices shape videos.
