A quick answer
Upload your video file and choose a frame. Original-image mode adds text without regenerating the person. AI modes can alter the scene and face. This workflow does not import a video from a YouTube URL.
Your video file
Work with footage you provide yourself.
Original frame with text
Keep the person in the selected frame.
A new visual idea
Choose an AI mode for a different scene.
How to approach it
Upload the video
Use footage you have permission to use.
Select a moment
Find a sharp frame with a clear subject.
Choose the mode
Use original-image mode or a generated draft.
Review the export
Check the person, message and readability.
Finding a useful frame
A still moment often gives a clearer image than fast movement. Look for open eyes and room for text.
The frame need not explain every chapter. It should set a reasonable expectation together with the title.
Keeping the original person
Original-image mode avoids generating a new person. The selected frame remains the image's foundation.
For a generated scene, inspect faces, clothing and objects closely. Keep the original as a reference.
Good to know
Can I paste a YouTube link?
This workflow needs your video file. A link does not replace the upload.
Will the person stay unchanged in every mode?
No. Generative modes can change their appearance. Use original-image mode when needed.