Create short AI dance videos from a text prompt and an optional reference photo. Bring your idea into focus with movement, lighting and camera direction—from a contemporary studio scene to a hip-hop rehearsal or a jazz performance.
Contemporary, hip-hop and jazz suggest different movement and visual choices. Explore these AI-created illustrations and suggested prompts to shape your own dance brief. The opening video is an AI-created dance showcase, edited from multiple clips.
A full-body contemporary dancer moves through a slow turn with open arm gestures in a quiet studio. Fixed wide shot, soft daylight, room around the hands and feet.
A full-body hip-hop dancer performs a short sequence of grounded steps and clear upper-body accents in a rehearsal space. Fixed wide shot, even lighting, uncluttered background.
A full-body jazz dancer performs a short phrase with extended arm lines and a controlled turn on a small stage. Fixed wide shot, warm stage lighting, space around the performer.
If you use a reference photo, choose one with a clearly visible face, an unobstructed body and room around the hands and feet. A simple background can make the subject easier to describe. Use a photo you have permission to upload. Use the reference photo to guide the dancer’s appearance, then describe the movement you want in your prompt.
Describe the performer, one main action, the setting and the camera view. For example: “A full-body dancer performs one controlled turn in a quiet studio, filmed in a fixed wide shot.” Start with a focused movement idea rather than several scene changes. This is prompt guidance, not a command that guarantees exact choreography.
Replace broad words such as “cinematic” with details that describe the image you want. “Warm side light with soft shadows” suggests a different direction from “cool daylight in a bright studio.” Keep the performer and movement description similar when exploring another lighting idea, while allowing for variation in each new result.
For full-body dance, describe a wide view with space above the head and around the hands and feet. Avoid framing that begins with important limbs already cut off. When you use exactly one reference image, its proportions determine the video aspect ratio, so prepare the image in the shape you want. Check the generated clip for movement leaving the frame.
Your prompt describes the performance. Workspace settings define the request. Neither guarantees that every movement or visual detail will match your idea.
Start with text alone or add still-image references. This workflow does not accept a dance video for motion transfer. Multiple images are visual references, not a sequence of animation keyframes.
Choose a duration from 3 to 12 seconds and a resolution of 480p, 720p or 1080p. Select the available aspect ratio separately; writing ‘vertical’ or ‘1080p’ in a prompt does not replace choosing the setting. With one reference image, the image determines the aspect ratio.
Watch the entire clip. Check hands and feet, balance, facial appearance, clothing changes and movement continuity. A strong first frame does not establish that the rest of the performance is usable. Exact choreography and consistent appearance are not guaranteed.
Credit cost depends on duration and resolution. A new generation is a separate request, so check the displayed cost before trying a revised prompt. A completed clip that does not suit your creative goal is not the same as a failed generation.
Explore character concepts, vertical social clips and different lighting moods. Use these AI-created illustrations to shape the performer, movement and setting in your own dance brief.
Describe a character’s clothing, setting and one main movement to explore a short performance concept. If you add a reference photo, use it as visual guidance rather than a promise of identity preservation. Review whether both the appearance and the movement suit your project.
For a vertical brief, choose 9:16 when the aspect-ratio control is available, or use a suitably proportioned single reference image. Leave space where you intend to add captions later. Captions, music and final editing need to be handled separately from this generation workflow.
Try describing the same movement idea in daylight, warm side light or a stage setting. Change one part of the brief to make your intention easier to compare, without expecting identical movement across outputs. Each new generation uses credits, so decide what you want to learn before requesting another clip.
Write the scene, confirm the request in your workspace, then review and save the result.
Write a prompt covering the performer, main movement, setting, camera view and lighting. Add a reference photo if it helps communicate the subject. Keep the scene focused and use images you have permission to upload.
Continue to the workspace and sign in to generate. Choose the duration, resolution and available aspect ratio, check your reference images, and review the displayed credit cost. Continuing from this page does not submit the generation by itself.
Watch the complete result and check the movement and visual details. Download the MP4 from the workspace if it meets your needs. Save it within 72 hours of becoming ready; the download is not permanent storage.
Write a description of the dance scene you want to create and optionally add still-image references. Continue to the workspace, sign in, and confirm the video settings and credit cost before generating. You do not need a reference photo to start. This workflow generates an interpretation of your prompt; it does not transfer choreography from an uploaded dance video.
Use JPEG, PNG or WebP images, up to 20 MiB each. You can add up to seven reference images at 480p or 720p, or one at 1080p. A clear, full-body reference is a useful starting point for full-body dance. Multiple references do not act as ordered movement frames. Only upload images you have permission to use.
Add a reference photo to guide the dancer’s visual appearance, then describe the movement you want in the prompt. Include the clothing, a main action and the camera view—for example, “a dancer in a black outfit performing a controlled turn, filmed in a full-body wide shot.” Keep the brief focused, then review the clip to choose the take that best fits your idea.
Choose a whole-number duration from 3 to 12 seconds and a resolution of 480p, 720p or 1080p. The aspect-ratio options are 1:1, 16:9, 9:16, 3:2 and 2:3 when the control is available. With exactly one reference image, the image determines the output aspect ratio instead of the separate selector. Check these settings in the workspace rather than relying on instructions written into the prompt.
New accounts receive 50 signup credits, valid for 30 days. Generation costs depend on duration and resolution. For example, a 5-second video currently costs 40 credits at 480p, 70 at 720p or 130 at 1080p. With an unused signup balance, that covers one 5-second 480p request, not every resolution. Check the displayed cost and your available balance before generating. Further attempts use additional credits. See Pricing for current plan details.
Download the MP4 from the result player in your workspace once the video is ready. Video downloads expire 72 hours after the result becomes ready, so save a copy promptly. Before sharing, review the full clip, obtain permission for identifiable people where needed, and check the rights to any images, music or other material you use. Review the Terms of Service and Privacy Policy; generating a clip does not automatically clear third-party rights.
Choose a performer, one main movement and a setting. Bring that brief to the prompt, review your options in the workspace, and decide what to create next.