Product

Existing YouTube Thumbnail Face Replacement

Edit an owned or authorized YouTube-style thumbnail by replacing only the man's face with an authorized identity reference while preserving the layout, pose, bass guitar, room background, lighting, high-contrast design, and existing text areas.

Output

16:9 ・ 4K

0 uses

0 views

Existing YouTube Thumbnail Face Replacement

16:9 ・ 4K

GPT Image 2

Create from this prompt

Copy the prompt, then open ChatGPT and paste it in.

Prompt Template

Template uses

Template views

Reference needed before use

Reference 1 - Existing ThumbnailRequired

Upload the owned or authorized current YouTube-style thumbnail whose layout, composition, pose, instrument, background, lighting, and text areas should be preserved.

Reference 0 - Authorized Identity ReferenceRequired

Upload the owned or authorized identity reference for the person whose face should replace the face in the existing thumbnail.

Replaceable fields

[Character Identity]the same authorized person shown in Reference 0

Use Reference 1 as the current owned or authorized YouTube-style thumbnail and Reference 0 as the authorized identity reference. Regenerate the thumbnail in a 16:9 composition while keeping the same layout and composition, but replace only the man's face so it matches [Character Identity]. Preserve the existing pose, pointing finger, bass guitar, room background, lighting, high-contrast thumbnail style, and all authorized existing Japanese typography. Accurately preserve five visible text or badge areas from Reference 1: the top blue banner, the large main headline, the bottom blue music or rhythm line, the yellow circular age badge, and the black square badge in the lower right. Do not redesign the thumbnail. Do not alter the existing layout. Do not add new logos, watermarks, text blocks, badges, props, characters, or background elements. The only correction should be replacing the face with the authorized identity from Reference 0 while keeping the rest of Reference 1 intact.

2. Open ChatGPT

One template = one prompt = one image