Text-to-video scenes
Describe the subject, action, camera movement, mood, and lighting. Seedance 2.5 converts that prompt into a short AI video.
Seedance 2.5 model page
Use Seedance 2.5 to create cinematic text-to-video and image-to-video clips with strong motion control, flexible aspect ratios, and 480p, 720p, or 1080p output.



Creative preview
Build scenes with a clear subject, camera move, lighting style, and reference images, then render a ready-to-share video.
Start creating
The video workspace below opens with Seedance 2.5 selected. Write a scene, choose duration, aspect ratio, and resolution, then generate a polished AI clip.
No result yet
Describe a scene and hit generate
Model highlights
Seedance 2.5 is best suited to short-form video concepts where movement, camera language, and visual consistency matter more than a single still frame.
Describe the subject, action, camera movement, mood, and lighting. Seedance 2.5 converts that prompt into a short AI video.
Add reference images to anchor characters, products, style, or composition while still controlling motion through the prompt.
Prompt for pans, pushes, handheld energy, slow motion, or product rotations to make the result feel intentionally directed.
Choose 5, 10, or 15 seconds, common portrait or landscape ratios, and output quality up to 1080p for social or campaign use.
Workflow
The strongest Seedance 2.5 prompts read like a miniature shot list. Keep the scene focused, then iterate one variable at a time.
Name the subject, action, location, lens feel, camera movement, and lighting. Avoid packing multiple scenes into one prompt.
Pick vertical, landscape, square, or classic social formats, then set 5, 10, or 15 seconds based on the channel.
Check framing, subject stability, and motion rhythm. Keep the best prompt and change only one detail for the next render.
Use cases
Create short vertical concepts for Reels, Shorts, TikTok, X posts, and paid social tests.
Generate product mood shots, launch teasers, and cinematic movement ideas before production.
Turn static concepts into moving references for directors, marketers, designers, and founders.
Explore characters, costumes, environments, and lighting directions before committing to a full sequence.
FAQ
Seedance 2.5 is an AI video generation workflow for creating short clips from text prompts and optional reference images. It is designed for cinematic motion, controllable framing, and social-ready output.
Yes. You can use reference images to guide the subject, style, or composition, then describe the motion and scene in the text prompt.
This Seedance 2.5 page supports 5-second, 10-second, and 15-second video generations.
The available Seedance 2.5 settings include 480p, 720p, and 1080p output options, depending on the selected configuration.
A good prompt describes one scene with subject, action, camera motion, lighting, style, and atmosphere. Specific direction usually gives stronger motion.
Yes. Seedance 2.5 is useful for product teasers, concept videos, social hooks, storyboard clips, and early campaign exploration.
Open the Seedance 2.5 workspace, write one focused shot, and generate a short cinematic clip in your browser.