Seedream Series

POST

tasks

Seedream Series

curl --request POST \
  --url https://api.qingbo.ai/v1/tasks \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "action": "<string>",
  "prompt": "<string>",
  "image_urls": [
    "<string>"
  ],
  "aspect_ratio": "<string>",
  "resolution": "<string>",
  "n": 123,
  "watermark": true,
  "optimize_prompt_options": {},
  "output_format": "<string>",
  "background": "<string>"
}
'

import requests

url = "https://api.qingbo.ai/v1/tasks"

payload = {
    "model": "<string>",
    "action": "<string>",
    "prompt": "<string>",
    "image_urls": ["<string>"],
    "aspect_ratio": "<string>",
    "resolution": "<string>",
    "n": 123,
    "watermark": True,
    "optimize_prompt_options": {},
    "output_format": "<string>",
    "background": "<string>"
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: '<string>',
    action: '<string>',
    prompt: '<string>',
    image_urls: ['<string>'],
    aspect_ratio: '<string>',
    resolution: '<string>',
    n: 123,
    watermark: true,
    optimize_prompt_options: {},
    output_format: '<string>',
    background: '<string>'
  })
};

fetch('https://api.qingbo.ai/v1/tasks', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.qingbo.ai/v1/tasks"

	payload := strings.NewReader("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"image_urls\": [\n    \"<string>\"\n  ],\n  \"aspect_ratio\": \"<string>\",\n  \"resolution\": \"<string>\",\n  \"n\": 123,\n  \"watermark\": true,\n  \"optimize_prompt_options\": {},\n  \"output_format\": \"<string>\",\n  \"background\": \"<string>\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.qingbo.ai/v1/tasks")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"image_urls\": [\n    \"<string>\"\n  ],\n  \"aspect_ratio\": \"<string>\",\n  \"resolution\": \"<string>\",\n  \"n\": 123,\n  \"watermark\": true,\n  \"optimize_prompt_options\": {},\n  \"output_format\": \"<string>\",\n  \"background\": \"<string>\"\n}")
  .asString();

Seedream is ByteDance’s line of image generation and editing models. This page covers the five models available today. They share one set of request parameters, one task endpoint and one billing model (a fixed price per successfully generated image), and differ in resolution tiers, aspect ratios and line-specific parameters — switching models means changing model and nothing else.

Quick start

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: seedream-demo-001" \
  -d '{
    "model": "seedream-4.0",
    "action": "generate",
    "prompt": "A celadon teacup on an oak table, natural window light, product photography",
    "aspect_ratio": "1:1",
    "resolution": "2k",
    "n": 1
  }'

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: seedream-edit-001" \
  -d '{
    "model": "seedream-5.0-pro",
    "action": "edit",
    "prompt": "Keep the subject and replace the background with a bamboo grove at dawn",
    "image_urls": ["https://cdn.example.com/source.png"],
    "aspect_ratio": "auto",
    "resolution": "2k",
    "n": 1
  }'

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedream-5.0-lite",
    "action": "generate",
    "prompt": "A minimalist product poster with generous white space and a fountain pen in the centre",
    "aspect_ratio": "3:4",
    "resolution": "4k",
    "output_format": "png",
    "watermark": false
  }'

curl -X POST https://api.qingbo.ai/v1/tasks \
  -H "Authorization: Bearer $WAVE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedream-5.0-flash",
    "action": "edit",
    "prompt": "Dress the person in image 1 in the coat from image 2, set against the street in image 3",
    "image_urls": [
      "https://cdn.example.com/person.jpg",
      "https://cdn.example.com/coat.jpg",
      "https://cdn.example.com/street.jpg"
    ],
    "resolution": "1.5k"
  }'

A successful submission returns a task_id. Retrieve the result with GET /v1/tasks/{task_id}, wait inline with Prefer: wait, or configure a webhook.

Request parameters

string

required

One of seedream-4.0, seedream-4.5, seedream-5.0-lite, seedream-5.0-pro or seedream-5.0-flash.

string

default:"generate"

generate: create an image from text
edit: rewrite a reference image; requires image_urls

string

required

The image description or editing instruction, in Chinese or English.

string[]

Reference image URLs. Required for edit: up to 10 on seedream-5.0-flash, one on the other models; generate does not accept references.

string

default:"1:1"

Framing. The 10 ratios shared by all five models: 1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 2:1, 1:2, 21:9.seedream-4.0 and seedream-4.5 also support 9:21.auto lets the model choose the ratio. seedream-5.0-pro and seedream-5.0-flash default to auto and accept it on both actions; the other three default to 1:1 and accept auto only on edit.

string

default:"2k"

Output resolution. The tiers differ per model:

seedream-4.0: 1k, 2k, 4k
seedream-4.5: 2k, 4k
seedream-5.0-lite: 2k, 3k, 4k
seedream-5.0-pro: 1k, 1.5k, 2k, defaulting to 1.5k
seedream-5.0-flash: 1k, 1.5k, 2k, defaulting to 1k

integer

default:"1"

Number of images to generate. seedream-4.0 and seedream-4.5 return one image for text-to-image; when editing, reference images + n may not exceed 15 (with one reference image, n is at most 14). seedream-5.0-lite accepts 1–15 for text-to-image, and the same reference images + n ≤ 15 rule applies when editing. seedream-5.0-pro and seedream-5.0-flash accept 1.With n above 1 the model works in multi-image mode: n is the maximum number of images for the request, the model may return fewer, and you are charged for the images actually returned.

boolean

default:"false"

Add a watermark to the output image.

string | object

Prompt optimisation mode, supported by seedream-4.0 and seedream-4.5 only, defaulting to standard.

standard: quality first, slower
fast: speed first

seedream-4.0 takes a string ("fast"); seedream-4.5 takes an object ({"mode": "fast"}).

string

default:"jpeg"

Output image format, jpeg or png. Supported by seedream-5.0-lite, seedream-5.0-pro and seedream-5.0-flash only.

string

default:"opaque"

Background: opaque or transparent. Supported by seedream-5.0-pro and seedream-5.0-flash only, and transparency requires image-to-image with a single alpha-channel input and output_format set to png.

See Submit a task for callback_url, callback_events, Prefer: wait, Idempotency-Key and the maximum-cost header. This series does not accept seed, quality, negative_prompt or mask_url. Any other parameter returns 400 and is not billed.

Limits

Condition	Limit
`seedream-4.0` / `seedream-4.5` / `seedream-5.0-lite`	A prompt is required; up to 15 images per request, and reference images count toward the 15.
`seedream-4.0` / `seedream-4.5`, `action: "generate"`	Text-to-image returns one image; multiple outputs need a reference image (edit).
`seedream-5.0-pro` / `seedream-5.0-flash`	A prompt is required and each request returns one image
`action: "generate"`	Takes no reference images
`action: "edit"`	Reference images are required: 1 to 10 on `seedream-5.0-flash`, exactly one on the other models
Reference images	Up to 10 on `seedream-5.0-flash`, one on the other models
`aspect_ratio: "auto"`	Available on `edit` only for `seedream-4.0`, `seedream-4.5` and `seedream-5.0-lite`
`optimize_prompt_options`	`seedream-4.0` (string) and `seedream-4.5` (object) only
`output_format` / `background`	`output_format` on the three 5.0 models; `background` on `seedream-5.0-pro` and `seedream-5.0-flash` only

Pricing

Billed at a fixed price per successfully generated image. seedream-4.0, seedream-4.5, seedream-5.0-lite and seedream-5.0-flash charge the same at every resolution they support; seedream-5.0-pro charges more at 2k than at 1k / 1.5k. Rates are in price_config on GET /v1/models and in the console’s Model Market. What a single call actually cost is the cost field on the task response — an integer quota at 500,000 quota = 1 USD, not dollars. generate and edit cost the same, reference images add no fee, and aspect ratio, output format and prompt length do not affect the price — the per-image price depends only on resolution, and a call costs the per-image price × the number of images actually returned. Quota for n images is frozen at submission; if multi-image mode returns fewer than n, the excess is refunded when the task completes. A resolution outside the tiers a model supports settles at its fallback price, equal to that model’s base tier. You are only billed for a successfully generated image. Failed and cancelled tasks, and tasks that return no usable image, are refunded in full.

Response

Completed task

{
  "task_id": "task-wave1789004688b950049238",
  "model": "seedream-4.0",
  "action": "generate",
  "status": "completed",
  "progress": "100%",
  "created_at": 1788754231,
  "completed_at": 1788754238,
  "result": {
    "images": [
      {
        "expires_at": 1789089508,
        "url": ["https://cdn.example.com/result.jpg"]
      }
    ]
  },
  "urls": {
    "get": "https://api.qingbo.ai/v1/tasks/task-wave1789004688b950049238",
    "cancel": "https://api.qingbo.ai/v1/tasks/task-wave1789004688b950049238/cancel"
  },
  "billing_status": "settled",
  "cost": 13500
}

Generated images come back in result.images, and expires_at is when that link stops working. See Query Task Status for the full field list.

Available models

Model ID	Resolutions	Aspect ratios	Line-specific parameters
`seedream-4.0`	`1k` `2k` `4k` (same price)	11 + `auto`	`optimize_prompt_options` (string)
`seedream-4.5`	`2k` `4k` (same price)	11 + `auto`	`optimize_prompt_options` (object)
`seedream-5.0-lite`	`2k` `3k` `4k` (same price)	10 + `auto`	`output_format`
`seedream-5.0-pro`	`1k` `1.5k` `2k` (`2k` costs more)	10 + `auto`	`output_format`, `background`
`seedream-5.0-flash`	`1k` `1.5k` `2k` (same price)	10 + `auto`	`output_format`, `background`

All five do text-to-image and reference editing and accept watermark. seedream-5.0-flash takes up to 10 reference images when editing, the other models one; seedream-4.0, seedream-4.5 and seedream-5.0-lite can return several images per request (see n).

The 10 shared aspect ratios: 1:1 4:3 3:4 16:9 9:16 3:2 2:3 2:1 1:2 21:9
seedream-4.0 and seedream-4.5 also support 9:21
auto lets the model pick the framing: seedream-4.0, seedream-4.5 and seedream-5.0-lite accept it only on edit; seedream-5.0-pro and seedream-5.0-flash accept it on both actions and default to it

Seedream Series

curl --request POST \
  --url https://api.qingbo.ai/v1/tasks \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "<string>",
  "action": "<string>",
  "prompt": "<string>",
  "image_urls": [
    "<string>"
  ],
  "aspect_ratio": "<string>",
  "resolution": "<string>",
  "n": 123,
  "watermark": true,
  "optimize_prompt_options": {},
  "output_format": "<string>",
  "background": "<string>"
}
'

import requests

url = "https://api.qingbo.ai/v1/tasks"

payload = {
    "model": "<string>",
    "action": "<string>",
    "prompt": "<string>",
    "image_urls": ["<string>"],
    "aspect_ratio": "<string>",
    "resolution": "<string>",
    "n": 123,
    "watermark": True,
    "optimize_prompt_options": {},
    "output_format": "<string>",
    "background": "<string>"
}
headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

const options = {
  method: 'POST',
  headers: {'Content-Type': 'application/json'},
  body: JSON.stringify({
    model: '<string>',
    action: '<string>',
    prompt: '<string>',
    image_urls: ['<string>'],
    aspect_ratio: '<string>',
    resolution: '<string>',
    n: 123,
    watermark: true,
    optimize_prompt_options: {},
    output_format: '<string>',
    background: '<string>'
  })
};

fetch('https://api.qingbo.ai/v1/tasks', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.qingbo.ai/v1/tasks"

	payload := strings.NewReader("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"image_urls\": [\n    \"<string>\"\n  ],\n  \"aspect_ratio\": \"<string>\",\n  \"resolution\": \"<string>\",\n  \"n\": 123,\n  \"watermark\": true,\n  \"optimize_prompt_options\": {},\n  \"output_format\": \"<string>\",\n  \"background\": \"<string>\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "application/json")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

HttpResponse<String> response = Unirest.post("https://api.qingbo.ai/v1/tasks")
  .header("Content-Type", "application/json")
  .body("{\n  \"model\": \"<string>\",\n  \"action\": \"<string>\",\n  \"prompt\": \"<string>\",\n  \"image_urls\": [\n    \"<string>\"\n  ],\n  \"aspect_ratio\": \"<string>\",\n  \"resolution\": \"<string>\",\n  \"n\": 123,\n  \"watermark\": true,\n  \"optimize_prompt_options\": {},\n  \"output_format\": \"<string>\",\n  \"background\": \"<string>\"\n}")
  .asString();

Text Generation

Image Generation

Video Generation

Audio

Tools & Embeddings

Task Management

Account & Usage

Quick start

Request parameters

Limits

Pricing

Response

Available models

​Quick start

​Request parameters

​Limits

​Pricing

​Response

​Available models

​Related

Quick start

Request parameters

Limits

Pricing

Response

Available models

Related