AI VIDEO EDITING

Replace a character in an existing video

Use an uploaded clip or a template as the starting point, then add an image of the replacement character. StirVid uses Higgsfield Genjutsu to edit the selected subject while using the source performance and scene as a guide.

Choose the right target

Identify the subject by position, clothing, or action: the dancer on the left, the singer in the striped shirt, or the skateboarder in the foreground. With several people in frame, specify each replacement separately.

Prepare your reference images

Choose a clear, well-lit character image with a visible face and body when possible. Additional views can show the same character from different angles; say so in the prompt instead of implying that each image is a different replacement.

A prompt to start with

Replace the dancer on the left in @video1 with the character in @image1. Keep the other dancers and background unchanged.

For multiple replacements, attach separate images and link each one: replace the person on the left with @image1 and the person on the right with @image2. Review the highlighted references before paying.

What to expect

Fast cuts, occlusion, tiny subjects, and major differences between the reference and source can make identity and motion less consistent. This is a generative edit, so unchanged parts of the scene may still vary.

Upload a 4–30-second MP4 and 1–8 reference images, then choose 480p or 720p output. Check your price before paying; no subscription or account is required. Confirmed failed, canceled, or provider-declined paid generations are refunded automatically.