Hop Hop Hop Dance AI: CapCut vs AI, other tools and prompts

Where a prompt drifts

• What you type: a general tool wants a long prompt and settings. MimicZip has no prompt box; you add two photos and choose the options.
• Two faces at two distances: models often make both people the same size, or give one person the other's face. Here the template fixes who is close and who is far.
• One hard cut: most models blend scenes. The swap at 4.2 seconds is built in.
• Cost: check each general tool's plans, because a two-person scene often needs several tries. MimicZip sells one-time packs from $8.99 with bonus previews and no subscription.

When a general tool is better

Use a general tool for your own choreography or full camera control. Use MimicZip when you want this one routine from your photos, fast. If you would rather write the prompt, see the guides for Kling, Veo and Seedance.

Copy this prompt

Image-to-video, 9:16, about 8 seconds. Fixed vertical phone camera (not visible in frame), wide-angle, no zoom, one hard cut on the beat.
Reference image 1 is the pointer and reference image 2 is the jumper; they are the only two characters. Keep each face, hair, skin tone and body shape in every second. Setting: a bright living room in soft daylight. No text or logos.
0 to 4.2 seconds: the pointer stands very close to the lens on the right of the frame, face large, grinning and swaying to the beat, one arm swung back over their shoulder pointing at the jumper. The jumper stands far back on the left, small, hopping to the beat, shaking out both arms, then dropping into a fast low squat and springing up.
At 4.2 seconds: a hard cut on the beat; the two have swapped places.
4.2 to 8 seconds: now the jumper is close on the right pointing back, and the pointer is far on the left hopping, shaking out their arms and dropping into a squat. End holding the squat.
Audio: instrumental upbeat bouncy beat with a clear kick. No voices, no counting, no lyrics.
Photorealistic, no text, no subtitles, no logos, no watermark.

Tips for Kling

• Attach the two photos as image 1 (pointer) and image 2 (jumper), full-body if you have them.
• Add the sound you like in TikTok afterwards. A text prompt cannot place the cut exactly on a beat.
• If the two look alike, add: the pointer and the jumper look clearly different from each other.
• Tools change often, so check the length and aspect ratio your tool supports.

Copy this prompt

Image-to-video, 9:16, about 8 seconds. Fixed vertical phone camera (not visible in frame), wide-angle, no zoom, one hard cut on the beat.
Reference image 1 is the pointer and reference image 2 is the jumper; they are the only two characters. Keep each face, hair, skin tone and body shape in every second. Setting: a bright living room in soft daylight. No text or logos.
0 to 4.2 seconds: the pointer stands very close to the lens on the right of the frame, face large, grinning and swaying to the beat, one arm swung back over their shoulder pointing at the jumper. The jumper stands far back on the left, small, hopping to the beat, shaking out both arms, then dropping into a fast low squat and springing up.
At 4.2 seconds: a hard cut on the beat; the two have swapped places.
4.2 to 8 seconds: now the jumper is close on the right pointing back, and the pointer is far on the left hopping, shaking out their arms and dropping into a squat. End holding the squat.
Audio: instrumental upbeat bouncy beat with a clear kick. No voices, no counting, no lyrics.
Photorealistic, no text, no subtitles, no logos, no watermark.

Tips for Veo

• Keep the clip close to 8 seconds, since Veo clips are short; if yours is shorter, drop the second round.
• Ask for a fixed, locked-off camera and one hard cut, or the model may move the camera and blend the swap.
• Say there are exactly two people and that nobody else appears.
• Leave out names, song titles and account names, and describe an instrumental.

Copy this prompt

Image-to-video, 9:16, about 8 seconds. Fixed vertical phone camera (not visible in frame), wide-angle, no zoom, one hard cut on the beat.
Reference image 1 is the pointer and reference image 2 is the jumper; they are the only two characters. Keep each face, hair, skin tone and body shape in every second. Setting: a bright living room in soft daylight. No text or logos.
0 to 4.2 seconds: the pointer stands very close to the lens on the right of the frame, face large, grinning and swaying to the beat, one arm swung back over their shoulder pointing at the jumper. The jumper stands far back on the left, small, hopping to the beat, shaking out both arms, then dropping into a fast low squat and springing up.
At 4.2 seconds: a hard cut on the beat; the two have swapped places.
4.2 to 8 seconds: now the jumper is close on the right pointing back, and the pointer is far on the left hopping, shaking out their arms and dropping into a squat. End holding the squat.
Audio: instrumental upbeat bouncy beat with a clear kick. No voices, no counting, no lyrics.
Photorealistic, no text, no subtitles, no logos, no watermark.

Tips for Seedance

• Attach your two photos as reference images 1 and 2 and, if the tool allows, a short reference clip for the camera only.
• Seedance follows second-by-second timing well: keep the time ranges in the prompt.
• Write the close and far positions explicitly (right and large, left and small) so the model keeps the distance.
• Ask for an instrumental beat and no voices, then add your sound in TikTok.