Human freestyle
One performer photo; create a short original shoulder-led scene.
Bulin 47 AI Street Freestyle Video Generator
Create a shoulder-led street performance from a photo, a prepared pet or two distinct portraits. Use an uploaded motion clip when you need reference-guided movement.
The street-freestyle effect places a lead performer beside a speaker partner and a relaxed spectator circle. Photo-based variations create a new short interpretation with shoulder accents, a small gesture and a held finish. The motion-reference variation instead follows an uploaded performance clip. Both approaches share one page so you can choose by the material you actually have.
Try this effect →Create your video
One performer photo; create a short original shoulder-led scene.
JPG, PNG or WebP, up to 10 MB. Use one clear subject per photo.
Seedance 2.5 · 720p · Reference-video seconds and output seconds are included in the quote. Images and audio use the existing pricing policy.
Read the English motion prompt
Use @Image1 as the sole lead identity, preserving face or markings, outfit and proportions. Make an original street-freestyle scene with a stable portable speaker held by an adult partner, an open spectator circle, visible hands and feet, and restrained background movement. The consenting adult turns toward the speaker, makes two alternating shoulder bounces, acknowledges the circle with a small hand gesture and holds the finish. Medium-wide continuous shot. Do not imitate a named artist’s face or voice.
Extra direction and the sound preference are added at submission. Model guidance does not guarantee frame-by-frame matching.
Estimated credits：1400 credits
Sign in to createBulin 47 AI Street Freestyle Video Generator · Prepare the inputs
Human and original-character variations use one clear identity image. A pet needs a prepared upright design with readable paws and shoulders. The two-person variation uses photo A for the lead and photo B for the speaker partner. Two identity photos are not first and last frames. For motion guidance, provide one identity image and an MP4 of the performance you want to follow.
Choose a variation
Pet freestyle
One prepared upright pet; animate readable shoulders and paws.
Original character
One mascot or character image; preserve silhouette and colors.
Your scene direction
One performer photo; describe the place, partner and gesture.
Bring your own moves
One identity photo and one motion clip; replace only the lead.
Performer and partner
Two identity photos: A performs, B holds the speaker.
How to make the video
Select the variation before uploading. In the photo-based workflow, choose the length and frame, then describe the place and gesture without inventing extra identities. For motion guidance, upload the clip and review its verified duration; the output must cover it. Use the same frame shape when you want to retain its composition. For a two-person performance, check both image positions before reviewing the quote. Submit once; reopening this page retrieves the saved task.
Review the result
Check the performer and speaker partner separately. Faces and outfits should remain assigned to the intended roles, and the speaker should stay stable in the partner’s hands. Look for merged limbs, moving feet outside the frame and sudden background motion. Pet and mascot variations should retain their original silhouette and markings.
For blended faces, supply one person per photo and choose the two-person variation. For poor motion matching, use the motion-reference variation and a clear short clip rather than adding more text. A pet snapshot should be prepared as an upright character first. Status polling only follows the existing task; it does not create another performance.
Common questions
Which variation follows my own moves?
Choose the motion-reference variation. Upload one identity photo and one MP4. The photo specifies the lead’s appearance; the video specifies movement, timing and framing.
Can the two uploaded people have different roles?
Yes. In the two-person variation, photo A is the lead performer and photo B is the speaker partner. Both are sent as separate identity references, without treating them as opening and closing frames.
Does this supply a particular artist’s song or voice?
No particular song or named voice is included. Photo-based performances request original sound when enabled. For a motion clip, the request prioritizes its soundtrack; you can provide a separate sound reference. Check the generated result before publishing.
The generated motion, likeness and soundtrack need to be reviewed before downloading.
Read this page as Markdown