• Every input mode in one endpoint — text-to-video, image-to-video, first/last frame, reference video, and reference audio.
• Real people, lip-synced — the face variants combine a reference video with audio to animate real people speaking: voiceovers, explainers, character animation in one call.
• Continuous chaining — set return_last_frame and feed the returned frame URL into the next call to chain clips without re-prompting.
• Per-second credit pricing — 1 credit = $0.001. Seedance 2.0 starts at $0.041/s; failed jobs are refunded automatically.

Seedance 2.5 will ship on the same request shape, so code written against Seedance 2.0 migrates by swapping the model id.

# 1. Submit a video task
curl https://reapi.ai/api/v1/videos/generations \
  -H "Authorization: Bearer $REAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedance-2.0",
    "prompt": "A barista pours latte art in slow motion, warm cafe lighting, cinematic 35mm look",
    "duration": 5,
    "resolution": "720p"
  }'
# → {"id": "task_..."}

# 2. Poll until status is "completed", then read output.video_urls
curl https://reapi.ai/api/v1/tasks/task_xxx \
  -H "Authorization: Bearer $REAPI_API_KEY"

Get a key from the reAPI dashboard — free credits on signup, no card required.

All four share one request shape: 4–15s clips, 480p–4K, prompts up to 4,000 characters, optional generated audio, aspect ratios from 21:9 to 9:16 plus adaptive. Try them in the playground without writing code.

The Seedance 2.5 page is live with positioning, pricing model, and use cases; the API itself is coming soon. Sign up now and your free credits will cover your first Seedance 2.5 generations at launch.

_{This is a developer resource for the Seedance API on reAPI, an independent API gateway. Not affiliated with, endorsed by, or sponsored by ByteDance. Model names belong to their respective owners.}