Skip to main content
seedance-pro-fast is seedance-pro on a faster runtime — same model weights and fidelity, just an optimized inference path that brings wall-clock down at a lower per-second price. The right pick when you’re iterating Seedance prompts in a loop and the wait between submit-and-result is your bottleneck. Pricing: $0.04 / second of generated footage — see the rate card. Failures don’t bill; per-second pricing applies to generated footage length, and the faster runtime is a runtime optimization, not a fidelity downgrade.

​
Protocols

ProtocolPathPurpose
OpenAI Video — submitPOST https://llm.bytespike.ai/v1/videos/generationsenqueues; returns task_id
OpenAI Video — pollGET https://llm.bytespike.ai/v1/videos/tasks/{task_id}returns status and result_url

​
Quickstart

TASK_ID=$(curl -s https://llm.bytespike.ai/v1/videos/generations \
  -H "Authorization: Bearer $BYTESPIKE_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "model": "seedance-pro-fast",
    "prompt": "Same prompt structure as seedance-pro, just faster runtime",
    "duration_seconds": 5,
    "size": "1280x720"
  }' | jq -r .task_id)
Poll loop matches the sora2 quickstart — model weights and surface are identical to seedance-pro, only inference time differs.

​
Capabilities

CapabilitySupported
Text-to-video✅
Image-to-video (with source_image)✅
duration_seconds 5 / 10✅
size 1280×720 / 1920×1080✅
Fidelity vs seedance-proidentical
Modalityvideo
Capability bucketvideo_generate

​
When to use

  • Prompt iteration loop — designer or marketer iterating wording / staging in tight loops; the runtime delta adds up over a session.
  • Demo / pitch playback — generate-and-show flows where the audience is watching the screen, wall-clock to first frame matters more than the per-second cost.
When not to use:

​
Next