Skip to content
Burning Bridges AI Video
Sign in

Burning Bridges AI video prompts

Copyable prompts for the Burning Bridges AI video: motion transfer, image-to-video, text-to-video, a vertical variant, a negative prompt and captions. Start with the honest part: a prompt alone will not give you the exact scene.

Read this first: what a prompt can and cannot do

Most of the viral Burning Bridges AI videos were not made from a text prompt alone. They used the party scene from Drake’s music video as a motion reference plus a photo of the new person, in a motion-transfer tool such as Higgsfield Genjutsu or Kling motion control. The tool copied the movement; the prompt, if there was one, was a short instruction on top.

So the best “prompt” for this trend is three inputs together: the clip, the photo and a sentence or two that tells the tool what to keep. A text prompt on its own can make a dim dinner party where someone covers their eyes and looks around. It will look like the scene. It will not be the scene: different room, different guests, different camera, different timing.

The prompts below are grouped by what you feed the tool. Use the first one if you have the clip. Use the others when you do not, and expect an approximation.

The prompts

1. Motion transfer instructionFor Higgsfield Genjutsu, Kling motion control and similar tools, with the party scene as the driving clip and a photo as the reference. Paste it into the instruction field.
Replace the lead in the video with the person in the reference image. Keep their face, hair, skin tone and build exactly as in the photo. Keep everything else from the video: the dim, chandelier-lit dinner party, the guests at the table, the camera movement and the timing. Keep the gesture: he walks through the room, holds one hand flat over his eyes and slowly scans the room. The face stays the same before, during and after the hand gesture. Natural skin texture, no extra people.
2. Image-to-video approximationFor image-to-video tools when you have a photo but no clip. It will look like the scene, not be the scene.
The person in the image walks slowly through an upscale dinner party at night. Dim warm light from crystal chandeliers, guests seated at a long table with candles and glassware, dark wood and deep shadows. Halfway through, he raises one hand flat over his eyes like a visor and scans the room from left to right, as if looking for someone. Calm, confident expression. Cinematic music-video look, shallow depth of field, slow tracking camera. Keep the face identical to the image.
3. Text-to-video approximationFor text-to-video tools with no photo and no clip, to test the look. The person will be invented.
Music-video shot, 10 seconds. A well-dressed man in a dark outfit walks through a crowded, dimly lit dinner party in a mansion. Crystal chandeliers, candlelight, warm amber tones, guests talking at a long table. He stops, raises one hand flat over his eyes and slowly scans the room, looking for someone. Smooth tracking camera at eye level, film grain, shallow depth of field, night.
4. Vertical 9:16 add-onAppend to any prompt above when you post to TikTok or Reels and the tool has no aspect-ratio setting.
Vertical 9:16 frame. Keep the lead centered from the chest up, with the raised hand and the eyes inside the frame for the whole clip. Chandeliers visible at the top of the frame, guests softly blurred behind.
5. Clean faces: negative promptFor tools with a negative prompt field. Remove any terms your tool rejects.
blurry face, distorted face, face morphing, changing identity, different person after the hand gesture, extra fingers, fused fingers, hand passing through face, second person making the gesture, plastic skin, over-smoothed skin, cartoon, watermark, text, logo, flicker, bright daylight, overexposed
6. Caption and hashtagsFor the post itself. Swap the first line for your own setup.
where she at?

Made with AI.
#burningbridges #drake #aivideo #burningbridgesai #wheresheat

How to adapt a prompt

  • Change one thing at a time. If you change the lighting, the outfit and the camera at once, you will not know which change fixed or broke the result.
  • Keep the gesture words. “Holds one hand flat over his eyes” and “scans the room” are what make it this meme. Without them it is just a party.
  • Keep the lighting words. “Dim”, “chandelier-lit”, “warm” and “night” hold the mood. Left alone, most models brighten the scene.
  • Describe the person only when there is no photo. With a reference image, extra description of the face fights the image; say “keep the face identical to the image” instead.
  • Ask for about 10 seconds. Longer clips give the model more time to drift, and the recognizable moment is short anyway.
  • Swap “he” for “she” or “they” to match the person. The gesture works the same.

Common failures

  • The face changes after the hand passes over it. Keep the line “the face stays the same before, during and after the hand gesture” and start the driving clip a second earlier.
  • The room comes out bright and generic. Repeat the lighting words and add “candlelight” and “deep shadows”.
  • Guests copy the gesture, or two people do it. Add “only the lead makes the gesture; the guests stay seated”.
  • The face looks waxy. Add “natural skin texture” and keep “plastic skin, over-smoothed skin” in the negative prompt.
  • Text or a logo appears in the frame. Put “text, logo, watermark” in the negative prompt.
  • The person does not look like the photo. That is usually the photo, not the prompt. The photo guide covers what to change.

The one-photo alternative

If you would rather skip prompts entirely, Burning Bridges AI Video takes one photo and uses the original party scene as the motion reference, so there is nothing to write or tune. You get a 720p video of about 10 seconds with the song audio, usually in a few minutes, for 1 credit. Packs start at $4.99. Upload your photo.

Prefer to do it yourself? How to make a Burning Bridges AI video walks through Genjutsu, Kling and CapCut step by step.

Frequently asked questions

Are these prompts free to copy?

Yes. Copy, edit and use them in any tool, no credit needed. The tools themselves may charge for each generation.

Will a prompt get me the exact Drake scene?

No. A text or image prompt makes a new video that resembles the scene: a different room, invented guests and different timing. The exact scene needs the original clip as the motion reference, in a motion-transfer tool or on Burning Bridges AI Video.

Can I paste a prompt into this site?

No. Burning Bridges AI Video has no prompt box. It takes one photo and uses the original party scene as the motion reference, so the result follows the real scene without any prompt.

Which aspect ratio should I use?

9:16 vertical for TikTok and Reels; 16:9 to match the music video or to post on X or YouTube. Generate in the ratio you will post in rather than cropping later, so the hand gesture stays in the frame.

Is this affiliated with Drake?

No. Burning Bridges AI Video is an unofficial fan tool and is not affiliated with, endorsed by or connected to Drake, OVO or Republic Records. The song and the original music video belong to their owners.

Related guides

Ready to see yourself at the party?

One photo, one credit, your own Burning Bridges AI video.

Make my video