BytePlus ModelArk Seedance 2.5: limits, tasks and how Sume maps
What BytePlus's ModelArk tutorial documents for Seedance 2.5 (task types, drafts, rate limits, retention) and which concepts exist in Sume's video API.
BytePlus ModelArk documents Seedance 2.5 with task types, a two-step draft flow, published rate limits and retention windows. Sume's POST /v1/videos has one job shape instead: no task-type field and no draft task, but the same seedance-2.5 model id.
Everything about ModelArk below is from BytePlus's Seedance 2.5 tutorial, read 2026-09-29.
What limits does ModelArk publish?
The tutorial lists rate limits by account type and retention windows for outputs.
| Item | Value |
|---|---|
| Individual account | 180 RPM, 3 concurrent tasks |
| Enterprise account | 600 RPM, 10 concurrent tasks |
| Generated video URL | Kept 24 hours |
| Task record | Kept 7 days |
| Output formats | mp4, mov |
What is the draft mode?
A draft is a 480p task, and the final render is 1080p only, billed as a second task; a draft task ID stays valid for 7 days. Sume has no draft flow. To preview cheaply on Sume, run the same prompt at resolution: "480p" as its own job.
How do ModelArk concepts map to Sume?
Sume's docs and API code define the right-hand column.
| ModelArk concept | On Sume |
|---|---|
| Task type: auto, edit, extend, reference | Not exposed; references go in input_references |
| 50 references | At most 12 input_references |
Ratio adaptive, duration -1 | Not in the seedance-2.5 list; send fixed values |
| Poll the task | Poll GET /v1/videos/{id} or pass callback_url |
Should I download output promptly?
Yes. The vendor keeps generated URLs for 24 hours; whatever Sume's retention is for unsigned_urls, copy the file into your own storage as soon as the job completes.
Sources
Related posts
More in Developers
- Seedream 5.0 censorship: what a blocked image looks like on Sume
Sume's Seedream calls run with the provider safety checker on, with no switch for it. What a refused generation returns, whether it is billed, and what to do.
- C# speech to text: transcribe audio files with HttpClient
Speech to text in C#: POST the audio file's URL with HttpClient, poll the job, then read the transcript and word timestamps from the JSON result.
- Put a still image above or beside a video with an API
POST /v1/timeline-1.0/compose stacks a still and a video into one MP4 (a half-banner), or overlays the still. Layout keys, limits and the flat price.
- Speech to text Korean API: send language_code or auto-detect?
Sume STT 1.0 takes an optional language_code such as ko; omit it and the language is auto-detected. When to pin Korean, when to omit it, and how TTS differs.
Written by Sume