Walking Away From Explosion Template

How the walking away from explosion template works, what stays fixed in the clip, and why one prepared template beats writing a video prompt from scratch.

What a walking away from explosion template is

A walking away from explosion template is a finished clip with two replaceable subjects. The road, the skyline, the camera move, the walk and the fireball are already locked in the template. Your job is to supply two photos: one of the person who should walk, and one of the car that should explode.

That is the difference between a template and a prompt. A prompt describes a scene in words and leaves the result to the model's interpretation, which is a poor fit for a fireball because a described explosion rarely comes back looking like a real one. A template ships the scene as a real clip, so the model copies the movement and the blast instead of inventing them. The instructions that remain are short and specific: replace this one person, replace this one car, keep the road, keep the timing.

What the template keeps fixed

The prepared clip carries all of the following:

  • The empty industrial road and the distant city skyline
  • The overcast colour grade and the natural light
  • The slow camera push and the framing of the walker
  • The single continuous take, with no cuts or transitions
  • The walk direction, the pace and the unbothered expression
  • The timing, size and position of the fireball
  • The original audio track

Because those are fixed, the result is repeatable. Render the same clip for five different cars and each one lands on the same beat with the same blast.

Why the template swaps two subjects

The template has exactly two things worth replacing: the person walking toward the camera and the vehicle that detonates behind them. Everything else, including the road and the camera, is the product.

That is why the generator asks for two photos. The first upload replaces the walker, so the face, hair and outfit come from your portrait. The second upload replaces the vehicle, so the body shape, silhouette, paint colour and trim come from your car photo. One person in frame and one vehicle in frame keeps each swap unambiguous.

What to upload

For the person slot, use a recent, front-facing photo with even lighting. Avoid sunglasses, heavy filters and anything that covers the face. You never need to match the original walker's stance, because the walk and the timing come from the template.

For the car slot, use a clear three-quarter or front photo of the whole vehicle in even light. Keep the paint colour readable, keep one car in frame, and avoid other vehicles behind it. The model takes the shape and finish from your photo, so a clean, well-lit shot of the full car gives the most faithful result.

What the template does not do

The template does not follow your photo's background, framing, camera angle or lighting. Those are explicitly excluded, so the result stays inside the prepared scene rather than turning into an animated version of your photo. The clip is an AI-generated, stylized fictional effect; nothing is filmed and nothing burns.

Ready to try it? The generator takes both photos and returns a finished clip.

Two photos, one prepared template, one finished clip.

Create Video