All templates

Kara Sevda AI Night Selfie Video Generator

Upload a photo

PNG, JPG or WebP · up to 10 MB

Choose a photo to generate a video.

Three faces in one night-selfie reaction

Upload one prepared trio photo for a Kara Sevda AI video. Animate a night-selfie lean, a jealous shoulder glance and a close 12-second love-triangle scene.

Kara Sevda uses a close night selfie to stage a small love-triangle reaction. The reference begins with a woman's face against a dark sky, then widens to show a suited man beside her and another man leaning over his shoulder. Their glances and changing distance make the joke clear without a large setting or an elaborate action sequence. The Kara Sevda direction uses the three adults already arranged in one image. Keep the lead near the phone, one companion beside them and the third behind the shoulder. A small lean toward the near companion and a restrained jealous glance can preserve the playful tension while keeping the faces distinct.

Best for

  • A television-inspired reaction scene
  • a friends' night-selfie joke or a close portrait with three clearly assigned roles. The format works through expression and proximity
  • so dark open sky and a little soft street light can be enough. Keep the composition intimate without letting faces overlap. The camera should feel handheld but calm
  • with each person's place obvious throughout the short exchange.

Arrange the lead and two companions together

  1. Choose your photo

    Upload one prepared image containing all three adults in the intended selfie arrangement. Show the lead closest to the lens, a companion beside them and a third person behind that companion's shoulder. Keep every face sharp and separated, with shoulders visible and room for the small lean. The preset does not add three independent identity-photo slots or guarantee the appearance of particular actors from a lone portrait.

  2. Generate with the template

    Keep all three existing faces in their assigned places and use only a small lean, glance and amused reaction.

  3. Review and download

    Check each face against its original place before following the expressions. The person behind should remain behind the same shoulder, and the lead should not exchange features with either companion during the lean. Watch cheek and shoulder contact, hair edges and jacket lapels in the close view. A subtle amused reaction is easier to read than a large turn that hides one of the three faces.

Guide the lean and jealous shoulder glance

Use this reference direction as a starting point in custom-prompt mode. The template runs fixed instructions; remixing is a separate workflow, not an edit to the preset.

Reference direction

Animate the three adults already arranged in the uploaded night-selfie image. Preserve each person's distinct face, hair, clothing, proportions and position. Keep the lead closest to the phone, the near companion beside them and the third adult behind the companion's shoulder. The lead makes a small affectionate lean toward the near companion, who turns slightly with an amused expression. The rear person leans forward just enough for a jealous sideways glance, then all settle into a readable close pose. Use gentle handheld camera drift and soft night lighting. Keep cheeks, shoulders and hair edges separate. One continuous scene; no added actors, identity exchange, forced kiss, captions, logos or rapid cuts.

Before you create

Can a single selfie choose the faces of both companions?

A Kara Sevda scene uses one prepared picture with all three people already present. That gives the scene a fixed arrangement and a separate visual identity for each role. It does not ask you to upload independent portraits for both companions or promise specific actor likenesses from an unrelated selfie. Prepare the trio first, then animate the small lean and jealous-shoulder reaction together.

Can I change the creative direction?

The preset uses fixed instructions. Copy the reference direction into custom-prompt mode to adapt it; this does not change the preset itself.

What video output should I expect?

The default length is 12s (reference ratio 540/960). The output follows your uploaded photo's aspect ratio and suits A television-inspired reaction scene, a friends' night-selfie joke or a close portrait with three clearly assigned roles. The format works through expression and proximity, so dark open sky and a little soft street light can be enough. Keep the composition intimate without letting faces overlap. The camera should feel handheld but calm, with each person's place obvious throughout the short exchange..

Create your Kara Sevda night-selfie video

Start with one prepared three-person selfie scene and build a close, expression-led love-triangle moment.