How to Make a Flash-Step (Dodge Step) AI Video
Follow a flash-step and dodge step dance tutorial using photo prompts. Learn how to choose a reference image, edit a transition and fix common AI video failures.
Want to make a flash-step video from a photo? Start with the reference clip below, then choose a full-body photo and describe one sideways move. The dance meme is also called Dodge Step; its Chinese name is 闪身步 (Shanshenbu), a move associated with Anhui Huagudeng. This guide covers photo animation and editing; the prompts have not yet been validated with generated clips.
Open the dedicated Flash-Step tool →
Original demonstration supplied for this tutorial, hosted in our cloud storage. This is not a ZNIX-generated result.
Turn the reference into a workable motion prompt
Watch the clip once for the overall idea, then pause around the move you want to recreate. Note the starting pose, travel direction and finishing pose. A reference clip helps you describe the movement; displaying it here does not feed its motion into the photo generator.
For the person prompt below, the sequence is still → shift weight → one short step to the right → feet together → pause. It deliberately asks for one move. To travel left, change every direction in the prompt and leave room on the left of your photo.
The cat and dog versions use a four-paw hop. Choose those for a pet remix; choose the person version if a human footwork sequence is essential. Neither option is a claim that the model reproduces every pose in the reference.
Change one variable per retry. If the subject moves out of frame, ask for a smaller step before changing the camera or background. If the face changes, keep the original clothes and scene and try a clearer photo. Save the prompt and settings next to each result so you can identify what helped.
Choose a photo for your flash-step video
Use one subject with an unobstructed face and the whole body visible. A cropped selfie gives the model too little information about legs and feet.
Leave space on the side the subject will move toward. Avoid groups, crossed limbs, busy backgrounds or a pet hidden behind furniture.
For a pet, choose a standing four-paw photo. A sitting close-up is a poor starting point for a lateral hop.
Make the first draft
- Open the dedicated Flash-Step tool. It fixes the original reference video for you.
- Select person, animal or fruit/object and upload one clear subject photo.
- When generation is available, review the live credit quote. Editing preserves the original aspect ratio and duration; the output is 720p and silent.
- Review the starting pose, footwork, sideways movement and finishing pause before downloading. The older prompts below are optional side-step experiments, not the dedicated tool’s recipe.
Pick a person, cat or dog prompt
Animate the single adult person in the reference photo into a short playful side-step video inspired by the flash-step (Shanshenbu) dance meme. Use a locked full-body camera, with both feet visible and clear space on the right. Begin still, gently shift weight to the left leg, take one quick small step to the right with a slight torso lean, bring the trailing foot alongside, then pause in a balanced stance. Show one clear lateral displacement, not continuous dancing or spinning. Preserve the face, hairstyle, outfit and body proportions. Keep feet in contact with the floor and movement physically coherent. One continuous shot, stable background, no cuts, no extra people, no extra limbs, no text or logos. Do not imitate a named person or existing recording.
Copying is free. Video generation uses credits; check the live quote before submitting. Review the result before spending credits on another attempt.
Edit a dodge step dance transition
These prompts request one continuous shot. For a travel transition, create or film a second shot with a similar subject size and the same movement direction, then cut at a matching moving pose.
Match the visual cut before adding music. Use a track you can use on your publishing platform, and place the beat at the step or landing. An image-to-video prompt does not guarantee the original soundtrack or exact beat synchronization.
Add captions such as “hearing about overtime” in your editor after generation; do not ask the model to render small text inside the clip.
Fix the common failures
| It dances but never moves sideways | Keep only one small lateral step or hop. Remove spins, complex dance sequences and moving-camera instructions. |
|---|---|
| The face or fur pattern changes | Use a clearer reference, one subject and one continuous shot. Remove clothing changes and background swaps from the same generation. |
| Extra paws or floating feet | Keep pets on four paws, show the floor and simplify the movement. Reject a broken clip instead of hiding it behind flashes. |
| It does not match the viral dance exactly | Exact timing and pose transfer generally needs a reference-motion workflow. This page does not offer or verify that capability. |
Check the cost of retries
The prompt is free; each generation or retry may consume credits. Prices depend on the available model, duration, resolution and audio settings. Compare the total credits spent with the number of usable clips, rather than judging the cost by a single attempt. No fixed price or success rate is promised.
Sources and related ideas
Reviewed on October 5, 2026. China News Service describes the cross-platform meme and everyday scenarios. Creator search results also show photo transitions and cat remixes; those examples are external references, not results generated or tested by ZNIX.
Gouxiong Duosuomao and Langzi Tiqiu are adjacent meme ideas. They require different motion descriptions; changing the title alone does not create a working template.
Frequently Asked Questions
Is Dodge Step the same as Flash-Step?
Can one photo make a Flash-step AI video?
Is the Flash-step AI template free?
Can I make my cat or dog do the dance?
Does this include the original music or a Jianying template?
Keep Reading
Topic hub: AI Video GeneratorAI Video Generator: The Complete Resource HubThe ZNIX editorial team explains AI models using linked primary sources, dated availability checks, and clearly labeled examples. Documentation summaries are distinguished from hands-on benchmarks.
Open the dedicated Flash-Step tool
Copying is free. Video generation uses credits; check the live quote before submitting. Review the result before spending credits on another attempt.
Open the dedicated Flash-Step tool