SeeDance 1.0 Mini
What is SeeDance 1.0 Mini?
SeeDance 1.0 Mini is the compact, lower-cost variant of ByteDance's first-generation SeeDance video line. It generates short clips of 5 or 10 seconds with three input modes: a text prompt, a first-frame plus optional last-frame pair, or a multi-frame keyframe chain with per-frame prompts and durations. Text-to-video and first_last_frame run at 720p or 1080p; multi-frame mode is locked to 720p. Multi-frame is the standout: it lets you stitch a sequence of keyframes into one continuous clip without a separate edit pass.
Key features of SeeDance 1.0 Mini
Four features cover what you actually use day to day.
Three input modes in one model
Text-to-video for prompt-only work, first-and-last-frame for bookended shots, and multi-frame for keyframe-driven sequences. One model ID, three call shapes.
Keyframe chaining via multi_frame
Pass a list of keyframe images, each with its own per-frame prompt and duration. Total clip length is the sum of segment durations. Useful for storyboard-driven work where you already have key art.
720p baseline, 1080p on text-to-video
Single-mode renders go to 720p or 1080p. Multi-frame is locked to 720p. Pick the resolution that matches the surface where the clip will play.
Six aspect ratios in text-to-video
Cover ultrawide, landscape, square, vertical, and tall mobile in one model. Framed and multi-frame modes inherit aspect ratio from the input images, so layouts stay consistent.
Best for
High-volume draft generation
Cheapest SeeDance 1.0 model. Good for prompt sweeps and batch jobs where queue time is acceptable.
Keyframe-driven storyboards
Multi-frame mode chains your keyframe images with per-segment prompts. Useful when you already have rough stills.
First-and-last-frame ad bumpers
Bookend the camera move with two stills and let the model interpolate. Common pattern for product reveals and brand stings.
Vertical and square social
9:16 and 1:1 ratios in text-to-video keep feed-native cuts in their target shape from the first render.
Internal review previews
Generate cheap previews before stepping up to SeeDance 1.0 Pro or 1.5 Pro for the final pass.
Use cases
Run a prompt sweep across ten variations of a hook before locking the wording. Build a 30-second sizzle reel by chaining six multi-frame keyframes in one call instead of stitching six separate clips. Storyboard a product reveal with first-and-last-frame mode using the SKU shot and a packaged shot. Produce vertical 9:16 cuts for social ads in the same render where the desktop 16:9 cut is generated, by re-running the same prompt with a different aspect ratio.
API examples
Call SeeDance 1.0 Mini from any language by POSTing to /v1/tasks. Full parameter docs live at docs.unifically.com/models/video/bytedance/seedance-1.0-mini.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "bytedance/seedance-1.0-mini",
"input": {
"prompt": "A butterfly landing on a flower in slow motion",
"mode": "text_to_video",
"aspect_ratio": "1:1",
"duration": 5,
"resolution": "1080p"
}
}'
Successful submission returns a task_id. Poll GET /v1/tasks/<task_id> or set a callback_url on the request to receive the finished result.
FAQs
People also ask
SeeDance 1.0 Mini is the compact variant of ByteDance's SeeDance 1.0 video line. It supports text-to-video, first-and-last-frame interpolation, and multi-frame keyframe chaining. Text-to-video and first_last_frame run at 720p or 1080p; multi-frame is locked to 720p. Clips are 5 or 10 seconds. It is the cheapest SeeDance 1.0 model on Unifically.
Three modes. text_to_video uses a prompt and aspect ratio. first_last_frame takes a required start frame and an optional end frame. multi_frame chains keyframe images with per-frame prompts and durations, locked to 720p.
5 or 10 seconds in text_to_video and first_last_frame modes. In multi_frame mode the total length is the sum of all per-frame durations.
21:9, 16:9, 4:3, 1:1, 3:4, and 9:16 in text_to_video. first_last_frame and multi_frame inherit aspect ratio from the input images.
No. Multi-frame is locked to 720p. If you need 1080p, run a text-to-video or first_last_frame call instead.
Mini supports an extra mode (multi_frame) that 1.0 Pro does not, and runs cheaper. 1.0 Pro outputs at 720p only and skips the multi-frame chain. Pick Mini for keyframe-driven workflows or volume drafts; pick Pro when you want the higher-quality single-shot render.