THE CREATION GUIDE · UPDATED OCTOBER 13, 2026

How to make a Burning Bridges AI video.

The idea is simple: a familiar party scene gets a new main character. With a fixed-template workflow, the reference performance and creative direction are prepared for you. Your part is choosing one good portrait.

Current availability: the studio shows whether generation is open and displays its credit cost. Photo previews are free. This version supports original virtual adults only, with one lead and up to five optional background characters.

1. Choose a portrait that gives the model a clear identity.

Use an original virtual portrait with one adult face, even light, and visible eyes. A front-facing or slightly turned pose is easier to interpret than a strong side profile. Avoid sunglasses, masks, motion blur, and heavy face filters. If the source is a group photo, choose a separate portrait rather than leaving the model to decide who is the lead.

Our photo preview accepts JPG, PNG, and WebP files up to 10 MB. The preview checks the file and image dimensions; it does not verify your identity or guarantee a generation result. See the photo preparation guide for more detail.

2. Use the preset scene.

The planned product uses one operator-supplied reference video. You do not need to find a clip, select a model, change frame settings, or write instructions. The reference establishes the setting, camera movement, and performance; your portrait supplies the lead character's identity.

Actual duration and export specifications will be shown before any payment when generation launches. A reference clip can be longer than a model's supported output, so we will confirm a tested duration rather than promise an unverified length.

3. Let the video model create the soundtrack.

The video model is instructed to use the reference video’s background music, rhythm and sound, keeping the performance synchronized. Audio output is enabled. You do not need to upload an audio file.

The result is delivered exactly as returned by the model, with no separate audio replacement. This direction does not guarantee exact reproduction of the reference recording; listen to the full result before sharing.

4. Review the full clip, with sound on.

• Check the face from beginning to end, particularly during a turn or when a hand crosses it.
• Look for distorted fingers, a changing cup, or a background person taking over the lead role.
• Listen for continuous music, appropriate volume, and a clean ending. Music quality and completeness depend on the model output.
• Check that the image and soundtrack play correctly in the downloaded file.

5. Share with the right context.

Use an original virtual portrait you have permission to use, review platform rules, and apply the AI-content label where required. Keep your downloaded master before editing or posting it. The goal is a playful new performance, not a claim that the synthetic event really happened.