Home Blog

Seedance 2.0 Mini: The Fast, Low-Cost Video API

Seedance 2.0 Mini cinematic video generation on EmpirioLabs

Jun 24, 2026

EmpirioLabs AI

Seedance 2.0 Mini is the fastest, lowest-cost tier in the Seedance 2.0 family on EmpirioLabs. It generates short cinematic clips with synchronized native audio, camera control, and stable motion from text, images, or video, at 480p and 720p. It is built for drafts, rapid iteration, and high-volume short-form production, and it runs on the same OpenAI-compatible API as the rest of the catalog.

Mini supports the full set of Seedance modes: text-to-video, image-to-video (first frame, or first and last frame), multimodal reference, video edit, and video extend. You pick the behavior with a single mode parameter, or let the model auto-detect it from your attachments.

Quickstart

Send a request to /v1/videos/generations with the model set to seedance-2-0-mini. Generation is asynchronous: you get a job_id back, then poll /v1/jobs/{job_id} until the video is ready.

curl https://api.empiriolabs.ai/v1/videos/generations \
  -H "Authorization: Bearer $EMPIRIOLABS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2-0-mini",
    "prompt": "A paper boat drifting down a rain-soaked street, cinematic, soft focus",
    "resolution": "720p",
    "duration": 5
  }'

To animate an image, add an image URL and set mode to i2v_first. To continue an existing clip, pass a video URL with mode set to extend. Native audio is on by default; turn it off with generate_audio: false.

Pricing

Seedance 2.0 Mini is pay-as-you-go, billed per second of video generated, with no subscription and no minimums. It is the most affordable tier in the Seedance 2.0 lineup, with separate rates for text or image input versus video input, and for 480p versus 720p. See the live rates on the Seedance 2.0 Mini model page and the pricing page.

Limits and tips

  • Mini outputs 480p and 720p. For 1080p or native 4K, use Seedance 2.0 Pro.
  • Clips run 4 to 15 seconds at 24fps.
  • Video inputs for reference, edit, and extend are capped at 15 seconds and normalized to a compatible format automatically.
  • For lifelike-face cohesion across inputs, generate the starting image with Seedream 5.0 Lite first, then pass it to Mini.

Try it now in the playground or read the full reference in the API docs.

Ready to use better endpoints?

Explore our models, or contact us about business inquiries, custom deployments, or anything else.