AI Video APIHalf-price Seedance iteration
Mini uses half the documented Seedance 2.0 rates for the same 480p and 720p resolution tiers, making it practical for prompt exploration, batch variants, and cost-sensitive production workloads.
Affordable multimodal video API
Seedance 2.0 Mini API
Generate high-volume 480p and 720p video at half the standard Seedance 2.0 price.
Seedance 2.0 Mini is the lightest, most cost-effective model in the Seedance 2.0 family. It generates 4–15 second clips at 480p or 720p from text prompts and optional image, video, and audio references, supports synchronized audio, and uses the same asynchronous task, polling, and callback workflow as the rest of AI Video API.
6 credits/s
Starting rate
4–15s
Duration
480p / 720p
Resolution
ByteDance
doubao-seedance-2.0-miniasync task
6 credits/s
Starting rate
4–15s
Duration
480p / 720p
Resolution
Model capabilities
Built for production API workflows, not one-off demos.
Use Seedance 2.0 Mini from the same platform surface as the rest of your video and image stack: API keys, credits, logs, webhooks, and docs stay consistent across providers.
Omni-reference generation
Combine a prompt with up to nine images, three videos, and three audio references. Use images for identity and composition, video for motion and camera rhythm, and audio for music or sound direction.
First and last frame control
Use one image as the first frame or two images as the first and last frames. In this dedicated mode, video and audio references are ignored and the request uses the no-video pricing tier.
Synchronized audio
Generate a synchronized soundtrack with the output clip. Audio references can guide the result when paired with at least one image or video reference; audio cannot be the only reference medium.
Reference motion and camera intent
Attach short public video URLs to guide movement, composition, pacing, or transitions. The combined reference-video duration can be up to 15 seconds and is included in reference-video billing.
Web-search-assisted prompts
Enable web_search when a generation needs current public information. The option uses the same request format and remains compatible with the asynchronous callback workflow.
Repeatable production pipelines
Control randomness with seed, request a clean last-frame image for connected clips, receive a task ID immediately, and use polling or callbacks to move completed output into downstream systems.
API workflow
Submit tasks, track progress, and return generated assets.
01
Describe the clip and references
Write the prompt and optionally attach public image, video, and audio URLs. Choose omni_reference for mixed media or first_and_last_frames for one- or two-frame control.
02
Choose efficient output settings
Select a 4–15 second duration, 480p or 720p resolution, and a fixed or adaptive aspect ratio. Enable synchronized audio, web search, seed, or last-frame output as needed.
03
Submit an asynchronous task
POST the request with model doubao-seedance-2.0-mini. The API validates the input, calculates the estimated credits, and immediately returns a task ID.
04
Receive output or a refund
Poll the task endpoint or provide callback_url. Successful tasks return hosted video URLs and optional last-frame output; failed generations automatically refund pre-charged credits.
POST /v1/videos/generations
curl -X POST https://api.aivideoapi.ai/v1/videos/generations \
-H "Authorization: Bearer sk-your-api-key" \
-H "Content-Type: application/json" \
-d '{
"model": "doubao-seedance-2.0-mini",
"callback_url": "https://your-server.com/webhook",
"input": {
"prompt": "Use the product image as the subject and follow the reference camera movement",
"generation_type": "omni_reference",
"image_urls": [{ "url": "https://example.com/product.jpg" }],
"video_urls": ["https://example.com/camera-motion.mp4"],
"audio_urls": ["https://example.com/music.mp3"],
"duration": 8,
"aspect_ratio": "16:9",
"resolution": "720p",
"generate_audio": true,
"return_last_frame": true
}
}'What teams build
High-volume short-video generation
Advertising and concept drafts
Social media variants
E-commerce product clips
Storyboard previews
Reference motion tests
Creator-tool integrations
Connected multi-clip sequences
Why choose Seedance 2.0 Mini
The Seedance 2.0 Mini API benefits that matter to developers.
Transparent reference billing
Text, image, and first/last-frame requests bill output duration. When video references are present, the input-video duration is detected and added to the output duration before applying the lower reference-video rate.
One multimodal request shape
Switch from Seedance 2.0 or Fast by changing the model ID while keeping the familiar prompt, image_urls, video_urls, audio_urls, callback, and generation-control fields.
Built for asynchronous delivery
The API separates submission from completion so high-volume systems can queue work, receive callback events, query task status, and process mirrored output without holding an HTTP request open.
Frequently asked questions
Answers about the Seedance 2.0 Mini API.
What is the Seedance 2.0 Mini API?
Seedance 2.0 Mini API is AI Video API's integration for doubao-seedance-2.0-mini, the lightest and most cost-effective Seedance 2.0 variant. It generates 4–15 second videos at 480p or 720p from text plus optional image, video, and audio context, returns a task ID immediately, and delivers hosted output through polling or callbacks.
How much does Seedance 2.0 Mini cost?
Without a video reference, Mini costs 9 credits per output second at 480p and 19 credits per output second at 720p. With video references, the rates are 6 and 12.5 credits per second respectively, applied to the combined input-video and output duration. A five-second 720p text-to-video request therefore costs 95 credits.
Which references and limits does Mini support?
Omni-reference mode supports up to nine images, three videos with a combined duration of at most 15 seconds, and three audio files with a combined duration of at most 15 seconds. Audio cannot be the only reference, so pair audio_urls with at least one image or video. Reference files must use public URLs that the service can retrieve.
Which durations, resolutions, and aspect ratios are available?
Seedance 2.0 Mini generates 4–15 second clips at 480p or 720p. It does not support 1080p or 4K. Supported aspect ratios are 16:9, 4:3, 1:1, 3:4, 9:16, 21:9, and adaptive, which selects a ratio from the prompt or reference media.
Can Seedance 2.0 Mini generate synchronized audio?
Yes. generate_audio defaults to true and creates synchronized audio with the video. You can also supply up to three audio references for music or sound direction, but audio references must be combined with at least one image or video and are ignored in first-and-last-frame mode.
How does first-and-last-frame generation work?
Set generation_type to first_and_last_frames and provide one image for the first frame or two images for the first and last frames. video_urls and audio_urls are ignored in this mode, and billing uses the without-video-reference rate. The frame workflow is useful for controlled transitions and connected visual sequences.
Does Mini support web search, seeds, and last-frame output?
Yes. Set web_search to true when the prompt needs current public information. Use seed to guide repeatability, with -1 selecting a random seed, and set return_last_frame to true to receive output.last_frame_url for chaining the result into another generation.
How do callbacks, task status, and refunds work?
The create endpoint returns a task ID immediately. Poll GET /v1/tasks/{taskId} or provide callback_url to receive completed or failed events. Successful responses include hosted output URLs and metadata; when generation fails, the pre-charged credits are automatically refunded.
Pricing and usage
Clear model options with shared credits.
480p without video reference
9 credits/s
Output duration × rate
480p with video reference
6 credits/s
Input video plus output seconds
720p without video reference
19 credits/s
Output duration × rate
720p with video reference
12.5 credits/s
Input video plus output seconds
Start building with Seedance 2.0 Mini in AI Video API.
Create one API key, use one credit balance, and switch between video and image models without provider-specific plumbing.