> ## Documentation Index
> Fetch the complete documentation index at: https://docs.imagine.art/llms.txt
> Use this file to discover all available pages before exploring further.
# Edit Video
> Transform existing videos using natural language commands — remove objects, change environments, adjust lighting, and more without traditional editing software.
Edit Video lets you modify an existing video by describing what you want to change in plain text. Rather than using a timeline editor or compositing tools, you write a sentence describing the edit and the AI makes precise, pixel-level modifications to the clip.
This mode is powered by Kling and supports a wide range of edits — from removing unwanted elements and swapping backgrounds to changing the weather, time of day, and visual style.
## What you can edit
Add, remove, or replace characters, props, or items anywhere in the scene. Remove distracting elements cleanly without visible artifacts.
Swap the background or relocate the scene entirely — from a city street to a forest, from an office to an open field.
Shift the lighting mood (sunset, dawn, overcast), change the time of day, or adjust colour temperature and shadow intensity.
Add atmospheric effects such as falling snow, rain, fog, or heat haze. Create dramatic environmental changes from a single description.
Apply a new visual style to the entire video — convert footage to look like a painting, sketch, or cinematic grade.
Alter the camera's perspective, tilt, or zoom to change the viewpoint of a scene without reshooting.
## How to edit a video
Navigate to [Video mode](https://www.imagine.art/video). From the Kling model options, select **Edit Video** to enter the editing interface.
Upload the video clip you want to edit.
**Input requirements:**
* Duration: 3–10 seconds
* Maximum file size: 200 MB
* Maximum resolution: 2K
Write a clear, specific natural language command describing the change you want to make.
**Example edit prompts:**
* `Remove all cars from the street and change the time of day to sunset`
* `Add snow falling in the background`
* `Replace the concrete wall with a dense forest`
* `Change the lighting to a warm golden hour glow`
* `Apply a vintage film grain style to the entire video`
Be direct and specific. Instead of "make it look more dramatic," describe the actual visual change: "add storm clouds overhead and reduce ambient light by half." Specific descriptions produce more predictable results.
You can upload up to 4 reference images or elements alongside the video to guide how new content should look. For example, if you want to replace a character's outfit, upload a reference image of the target clothing style.
When a video is present, you can upload up to **4 images or elements combined**. Without a video, you can upload up to **7 images or elements**.
Choose the video duration, resolution, and output format for the edited result.
Click **Generate**. The AI applies your described edits to the uploaded video while preserving the elements you haven't asked to change, maintaining subject consistency throughout the clip.
## Supported input media
| Media type | Specification |
| --------------------------- | ----------------------------------- |
| Video quantity | 1 video per generation |
| Video duration | 3–10 seconds |
| Video max file size | 200 MB |
| Video max resolution | 2K |
| Reference images | Up to 4 images (with video present) |
| Reference images (no video) | Up to 7 images or elements |
## What to do next
Add 5 seconds of new content to the end of your edited video.
Generate a new video from scratch using reference images to guide appearance.
Add talking-avatar lip movements to a video using audio or text.
See credit costs for Edit Video generations.