1
Choose a people-focused clip
Open a local video with a clearly visible person and check the source preview.
Browser-local video matting
Remove Video Background
Keep the person and change the scene. Blur the original background or replace it with a color or your own image, then export locally. Free, without a green screen, account or added watermark.
Local processing
CPU fallback
How to change a video background
2
Generate the person mask
Start removal and wait for the model and frame processing. Keep the tab open.
3
Choose and arrange the background
Try blur, a color or an image. For color/image modes, move or resize the person; image backgrounds also have Fit/Fill controls.
4
Export and check the file
Review the silent composite preview, export, then open the downloaded MP4 or WebM to check picture and sound. Changing background or placement requires another export.
Video inputs and limits
One video up to two minutes and 700 MiB. Codec support and available memory still matter. Processing requires supported browser video decoding and encoding; a file extension alone is not enough.
MP4WebM
Example: replace a presenter’s room
For a 20-second introduction, generate the mask once and try a plain color before choosing an office photo. Play the section where the presenter raises a hand to check the moving outline. In image mode, select the person or background layer to adjust its size and position independently.
Fit shows the whole background image with possible white margins; Fill covers the frame and may crop it. Blur keeps the person in the original placement and disables layout changes. Changing backgrounds reuses the mask, but does not repair missing foreground details.
Trim a longer recording, then add subtitles after exporting. Browse all free tools.
Video background removal questions
Is this background remover free?
Yes. No account or credits are required, and no DojoClip watermark is added. Save the finished file before leaving.
Do I need a green screen?
No. The model estimates a person mask from ordinary footage. Steady lighting and contrast between the person and surroundings help; it is not a tool for selecting and removing a specific screen color.
Is my video uploaded?
No. The model downloads to your browser and may be cached for reuse. Your video and chosen background image are processed on the device. The first model download needs a connection.
Can I keep a product or pet instead?
This tool is designed for people, without a selector for arbitrary objects. It does not erase an object and reconstruct the scene behind it. Objects held near the body may not remain cleanly separated.
Can I download transparent video?
No. The exported person is combined with the selected background into one opaque video. A WebM result is opaque too. The reusable mask is for this session’s composition, not an alpha-video download.
Why do hair or moving edges look soft?
Motion blur, low contrast and compression can confuse the estimated mask. There is no manual edge-painting tool. Check several moments, especially hands and hair. Output is re-encoded and limited to a 1920-pixel longest side and about two megapixels overall.
Will sound be included?
MP4 uses supported H.264/AAC; WebM uses VP9/Opus. Source audio must be decodable and encodable in a compatible combination, otherwise export can fail. Silent source videos also work. Review the downloaded picture and sound.