Creative Templates · 3 min read

How to make a two-character fateful gaze AI video

Recast both leads in a silent arena stare-down using two character images and one ready-made reference.

Try the two-character template

Updated 2026-10-14

The silent Fateful Gaze reference: balcony, crowd-wide shot and dance. Your two images recast its two leads; this preview is not a result made from your photos.

The short answer

Open the Fateful Gaze template, then upload two different character images you own or may use. Image 1 replaces the person first seen at the balcony; image 2 replaces the person on the arena floor and in the dance close-ups. The 15-second silent reference guides the shot order, gaze, crowd and movement, but generated details may vary. There is no prompt or soundtrack to copy. At 700 credits per output second, this template costs 15 × 700 = 10,500 credits, shown before submission. This two-character workflow is currently available for this template only; uploading your own video still replaces one main subject.

Step by step

Open the Fateful Gaze template
Watch the full 15-second silent reference before you choose characters. The first role appears at the balcony, then in the crowd-wide shot; the second is on the arena floor and in the later dance close-ups. You do not need to find or upload a trending show clip.
Put each character in the correct slot
Upload one clear JPG, PNG or WebP for each role, up to 10MB per image. Use two distinct adult characters with visible faces and different hair or clothing. Slot 1 is the balcony role; slot 2 is the arena-floor role. You must have rights to use both images.
Review the price and submit once
The reference is silent, so this result is silent too. The quote is 15 × 700 = 10,500 credits at the current rate. Confirm the rights statements, then Generate. The job appears in Library while it processes; you do not need to press Generate again. A technical failure releases reserved credits.
Check both identities before sharing
Watch the balcony, the crowd-wide cut and the dance close-ups. Check that role 1 and role 2 do not trade faces or clothes. If you add music later, use a track you have permission to publish; the template does not include the trend's original soundtrack.

Try the two-character template

What the reference controls

Two roles

The first image maps to the balcony lead; the second maps to the arena-floor lead. Review identity consistency across every cut.

Scene and motion

The reference guides the balcony-to-arena cut, crowded atmosphere, gaze and dance timing. It is not a frame-perfect guarantee.

Audio

This public reference contains no audio track. Add only sound you have the rights to after generation.

Scope

Two-character replacement is available for this approved template. Custom uploaded reference videos currently support one main subject.

What actually makes a difference

Make the two faces easy to distinguish

Different hair, clothing and silhouettes help the model keep the roles apart through the wide shot and close-ups.

Choose clean individual portraits

One adult per image, facing camera, evenly lit, without another person or a mask obscuring the face.

Review the distant crowd shot

Faces are smallest there, so check role assignment after the balcony cut instead of judging only the close-ups.

Keep the cast and score yours

Use original or properly licensed character images and audio. AI generation does not grant rights to someone else's show characters, actor likenesses or music.

Questions

Can I replace both people in any video I upload?

Not yet. Two image slots are available for the Fateful Gaze template. Custom uploaded videos currently replace one main subject.

Which image goes in which slot?

Image 1 is the person first seen at the balcony. Image 2 is the person on the arena floor and in the later dance close-ups.

Does the template include the viral music?

No. The public reference and generated result are silent. Add music you have permission to use in your editor afterwards.

What does one result cost?

The current rate is 700 credits per output second. This 15-second template is 10,500 credits, displayed before you submit.

Will both faces stay identical in every frame?

That is the goal, not a guarantee. Review the wide shot and each close-up for role swaps or visual drift before publishing.

Where do I find the finished video?

The job appears in Library as soon as it is submitted. If a technical failure occurs, reserved credits are released.

Ready to try it?

Recast both leads in a silent arena stare-down using two character images and one ready-made reference.

Try the two-character template