Home › Prompt

Burning Bridges AI Prompt Templates

4 templates · copy & paste · tested phrasing from the community

These are the prompt templates for the DIY route (the workflow is in our step-by-step guide). On our own generator you never need a prompt — the template handles it — but if you're using Higgsfield & Genjutsu yourself, these are the phrasings that work.

Start with Template 1. If you hit a "restricted content" error, it's usually the artist's name in the prompt — the no-name phrasing in Template 1 avoids it entirely.

Template 1 · No-name version (recommended)

Works the same as the original phrasing but never trips content filters, because no artist is named.

Replace the lead performer in this video with the person from my reference
photos. Keep everything else exactly the same: movements, gestures, timing,
camera motion, lighting, background, and other people in the scene. Match my
face, skin tone, hair, and body proportions accurately from the reference
images. Keep my face consistent in every frame, follow the original head
movements and lip movements naturally, and match the lighting and shadows of
each shot.

Template 2 · Original version (for reference)

The version that first circulated in the community. Same body as Template 1 but names the artist — some tools accept it, many return a restricted-content error. Included so you recognize it, not so you use it.

Replace Drake in this video with the person from my reference photos. Keep
everything else exactly the same: movements, gestures, timing, camera motion,
lighting, background, and other people in the scene. Match my face, skin tone,
hair, and body proportions accurately from the reference images. Keep my face
consistent in every frame, follow the original head movements and lip movements
naturally, and match the lighting and shadows of each shot.

Template 3 · Pet version

Yes, it works with pets — one of the funniest variants and oddly underused.

Replace the lead performer with the pet in my reference photos. Keep everything
else exactly the same: movements, timing, camera motion, lighting, background,
and other people in the scene. Match my pet's fur color, markings, face shape
and body proportions accurately from the reference images. Keep the pet's face
consistent in every frame.

Template 4 · Duo version

For two people in one clip. Reference photos should be labeled left and right.

Replace two people in this video with the two people from my reference photos.
Keep the left and right positions the same. Keep everything else exactly the
same: movements, gestures, timing, camera motion, lighting, background and the
other people in the scene.

Three things that matter more than the prompt

• The photo. Clear, front-facing, one person. A photo where the back of your outfit is also visible (a 360-style shot) noticeably improves the result.
• The source clip. Pick a short, well-lit segment — dark frames confuse the motion transfer.
• Expectations. Results are stylized and vary between attempts; plan on a retry or two.

Or skip all four templates — the one-click generator has the scene and prompt built in. Upload a photo, download a clip.

FAQ