YM2API Developer docs
Skip to content
Docs / Seedream 5.0 ProDownload Markdown ↓

Seedream 5.0 Pro · Images API

Group #186「【Seedream 生图】Seedream 5.0 Pro · 按量」, model seedream-5-pro, OpenAI Images format: POST https://yyds.seqnode.com/v1/images/generations; the response links to our image domain.

Price

Billed by tokens: $3.00 / 1M tokens for input and output, group rate 1. Measured usage per image (the model plaza is authoritative):

sizeOutputTokensPer image
omit size (default 1.5K)1536 × 1536≈ 76,500≈ $0.23
"2K"2048 × 2048≈ 153,000≈ $0.46

Parameters

FieldRequiredNotes
modelrequiredseedream-5-pro
promptrequiredImage description
sizeoptionalOmit for 1.5K; "2K" gives 2048 × 2048. Higher resolution uses more tokens
noptionalMust be 1: one image per request

Examples

Default 1.5K

curl --fail-with-body 'https://yyds.seqnode.com/v1/images/generations' \
  -H 'Authorization: Bearer YOUR_GROUP_186_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "seedream-5-pro",
  "prompt": "A rainy alley in a water town, wet stone path reflecting light, ink-wash style",
  "n": 1
}'

2K

curl --fail-with-body 'https://yyds.seqnode.com/v1/images/generations' \
  -H 'Authorization: Bearer YOUR_GROUP_186_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "seedream-5-pro",
  "prompt": "A rainy alley in a water town, wet stone path reflecting light, ink-wash style",
  "size": "2K",
  "n": 1
}'

Python · request and save

import requests

r = requests.post("https://yyds.seqnode.com/v1/images/generations",
    headers={"Authorization": "Bearer YOUR_GROUP_186_API_KEY"},
    json={"model": "seedream-5-pro", "prompt": "A rainy alley in a water town, ink-wash style", "size": "2K", "n": 1},
    timeout=300)
r.raise_for_status()
url = r.json()["data"][0]["url"]
open("seedream.png", "wb").write(requests.get(url, timeout=120).content)   # links expire after about 1 hour
print(url)

Response

On success data[0].url links to our image domain and expires after about 1 hour; download it right away. 2K usually takes about a minute, so allow a 5-minute timeout.

Common errors

StatusCauseWhat to do
400n is not 1, or wrong modelOne image per request; the model is seedream-5-pro
401Wrong key, or not a #186 keyUse a key of group #186
429 / 5xxUpstream busy or temporary failureRetry later; check usage first so you do not generate twice

Verification

2026-10-27, tested through this site: no size returned 1536 × 1536 (76,500 tokens); "2K" returned 2048 × 2048 (153,000 tokens, about 71 s). The former #141 Seedream / ZImage group was retired on 2026-10-30.

Back to the API docs · Group #186 in the model plaza