Koor AI

Practical guide · check one check two

Plan a Check One Check Two-inspired freestyle clip

A Check One Check Two-inspired freestyle edit can combine a confident close performance with a second appearance of the same character. Plan the doubled composition and the movement as separate jobs. This recipe builds an original night-time performance and an edit point for a clone reveal; it does not import another creator’s recording or soundtrack.

Open the photo-to-video studio

Generated demo and test notes

check one check two — generated workflow demo

Kling 3.0 Standard · 5.042 seconds · 828 × 554 · Generated Oct 22, 2026 (UTC+8)

A generated night-time performance is shown in two panels, with the right panel mirrored to demonstrate a simple clone-edit layout.

Both panels come from one take. The model did not create a second person; the split-screen edit does not reproduce the original song, voice or a seamless same-scene clone. Starting images were generated with FLUX.1 Schnell; this web copy is compressed and silent.

Input, prompt and method

View the fictional AI-generated starting image

Animate the uploaded adult in an original night-time street-performance scene. Preserve their face, age and outfit. They lean slightly toward the steady waist-up camera and make a short sequence of restrained head bobs and natural shoulder movements with a confident expression. Keep the face clear, eyes coherent and hands separate. No specific spoken words, no celebrity imitation, no captions or logos. End with a calm pause that can be cut to a separately prepared doubled-character shot.

This clip is AI generated. It is one example, not a success-rate measurement. Model, input, output and editing details are recorded above; no original social-platform soundtrack or template is supplied.

Choose an original prompt direction

Choose Use this look to load the prompt into the studio below. Select Image to Video and upload your authorized starting image before submitting. Review the selected model and credit quote first.

  • Solo Freestyle Setup

    Build one readable performance before adding a second appearance.

    Photo-to-video shot prompt
    Animate the uploaded adult in an original night-time street-performance scene. Preserve their face, age and outfit. They lean slightly toward the steady waist-up camera and make a short sequence of restrained head bobs and natural shoulder movements with a confident expression. Keep the face clear, eyes coherent and hands separate. No specific spoken words, no celebrity imitation, no captions or logos. End with a calm pause that can be cut to a separately prepared doubled-character shot.

  • Soft Streetlight Performance

    Use a single soft streetlamp as the dominant light, a plain jacket and an uncluttered background. Keep the short head movement relaxed.

    Photo-to-video shot prompt
    Animate the uploaded adult in an original night-time street-performance scene. Preserve their face, age and outfit. They lean slightly toward the steady waist-up camera and make a short sequence of restrained head bobs and natural shoulder movements with a confident expression. Keep the face clear, eyes coherent and hands separate. No specific spoken words, no celebrity imitation, no captions or logos. End with a calm pause that can be cut to a separately prepared doubled-character shot. Use a single soft streetlamp as the dominant light, a plain jacket and an uncluttered background. Keep the short head movement relaxed.

  • Doubled Frame Animation

    For this variant, the starting image already contains two separate appearances of the same adult. Keep both identities consistent and both figures separated; animate only a small coordinated head-and-shoulder movement.

    Photo-to-video shot prompt
    Animate the uploaded starting frame showing two clearly separated appearances of the same adult in a simple night-time street scene. Keep both faces, jackets, proportions and positions consistent. Both characters make a short restrained head-and-shoulder movement with confident expressions, then pause. Steady camera, neutral fill on both faces, no crossing limbs, no disappearance or new duplication, no specific lyrics, text or logos.

These are editable prompt recipes, not a gallery of verified generated results. Review the actual output from the model you choose.

Direct your own video

The studio uses the model shown in its selector. For this photo workflow choose Kling 3.0 Image to Video, add your first frame and start with a short test.

Add inputsAttach images, videos, or audio that are part of what you want to create.
0/2

JPG, PNG, WebP · 20 MB max

0/3

MP4, WebM, MOV · 100 MB max

0/3

MP3, WAV, M4A, FLAC · 50 MB max

Just add the material you need. The generator will use it automatically.

0 / 5000
Start with Kling 4.0 and preview the credit cost before generating. 36 welcome credits cover two 3-second Kling 4.0 videos at 720p, or three with Kling 3.0.
Choose modelCompatible models · settings update automatically

Only live routes are selectable. Your prompt stays in place when you switch.

Aspect ratio
Duration
Resolution
Prompt detail
This generation−15creditsCost guide

In short

  • Build one readable performance before adding a second appearance.
  • Use your own authorized character and original street-scene styling.
  • Prepare the doubled frame explicitly rather than expecting an automatic clone switch.
  • Add permitted sound and exact edit timing after generation.

Prepare the original and doubled compositions

Use a clear authorized adult reference with a visible face and upper body. A plain jacket and a softly lit night street create a performance setting without requiring a celebrity likeness or somebody else’s footage.

If you want two appearances in the same shot, compose the two versions in a starting image first. Keep their faces readable and their bodies separate. The photo-to-video model animates that first frame; it does not automatically know where a clone should enter halfway through the clip.

Give the edit three simple beats

  1. Begin close to the lens Hold a confident look with a little room beside the face for the later second appearance.
  2. Animate a short performance Use restrained head and shoulder movement. A silent performance avoids pretending that generic image animation can reproduce a particular rap recording.
  3. Reveal the second appearance Cut to a matching doubled composition or a second angle. Match the light and camera height, then align the edit with a permitted sound cue.

Inspect likeness and edit timing

Check both versions of the character for changing eye shape, inconsistent clothing and merged outlines. A short usable performance followed by a controlled edit is preferable to a long take that changes identity.

The recipe is an original interpretation, not an exact lip-sync or an imported template. Review generated audio independently if you enable it. Use your editor to control any recording, captions and the exact point where the second appearance is revealed.

Frequently asked questions

Does this reproduce the original freestyle recording?

No. The recipe creates an original performance concept. It does not supply or license another creator’s voice or recording.

Can I get a clone from one photo automatically?

Prepare a starting composition with the appearances you want, or cut between a solo shot and a separately composed doubled shot. This page does not include a fixed clone template.

Will the mouth match my chosen audio?

Exact lip-sync is not promised by this image-animation recipe. Align a permitted track in your editor and judge whether the visual result suits it.

Sources

Updated 2026-10-22. Koor AI is an independent video studio, not affiliated with or endorsed by Kuaishou or Kling AI. Kling model names belong to their respective owners. Each generation uses the model selected in the workspace.