Doubao Seedream 4.0 Image Generation

ByteDance's Doubao Seedream 4.0 model, supporting 1K/2K/4K multi-resolution generation, multi-image mode, and reference image input, with excellent performance in Chinese typography and Eastern aesthetics.

Try it Online

Test in Playground ↗ · View Models ↗

Quick Start

from openai import OpenAI

client = OpenAI(
    base_url="https://open.dieyuyun.com/v1",
    api_key="YOUR_API_KEY"
)

response = client.images.generate(
    model="doubao-seedream-4.0",
    prompt="An orange tabby cat sitting on a windowsill watching the snow, watercolor style",
    size="4096x4096",
    n=1
)

print(response.data[0].url)

Endpoint

Item	Value
Method	POST
Path	/v1/images/generations
Base URL	https://open.dieyuyun.com
Protocol	OpenAI Images API

Authentication

Bearer Token: Authorization: Bearer YOUR_API_KEY

Standard Request Fields

Field	Type	Required	Default	Description
model	string	Yes	—	Model identifier: doubao-seedream-4.0
prompt	string	Yes	—	Image description text, supports Chinese and English, recommended under 300 Chinese characters
size	string	No	4096x4096	Image size (width x height in pixels)
n	integer	No	1	Number of images to generate (max 15). When >1, multi-image mode is enabled
response_format	string	No	url	Response format: url or b64_json

Supported Sizes

The following recommended pixel values are available for each resolution tier:

1K Resolution

Size	Aspect Ratio
1024x1024	1:1
1152x864	4:3
864x1152	3:4
1312x736	16:9
736x1312	9:16
1248x832	3:2
832x1248	2:3
1568x672	21:9

2K Resolution

Size	Aspect Ratio
2048x2048	1:1
2304x1728	4:3
1728x2304	3:4
2848x1600	16:9
1600x2848	9:16
2496x1664	3:2
1664x2496	2:3
3136x1344	21:9

4K Resolution (Default)

Size	Aspect Ratio
4096x4096	1:1
4704x3520	4:3
3520x4704	3:4
5504x3040	16:9
3040x5504	9:16
4992x3328	3:2
3328x4992	2:3
6240x2656	21:9

Extended Parameters

Field	Type	Required	Default	Description
image	string/array	No	—	Reference image URLs or Base64, up to 14
watermark	boolean	No	false	Whether to add an AI-generated watermark
sequential_image_generation	string	No	disabled	auto for multi-image mode / disabled for single image

Request Examples

cURL

curl https://open.dieyuyun.com/v1/images/generations \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "doubao-seedream-4.0",
    "prompt": "An orange tabby cat sitting on a windowsill watching the snow, watercolor style",
    "size": "4096x4096",
    "n": 1
  }'

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://open.dieyuyun.com/v1",
    api_key="YOUR_API_KEY"
)

response = client.images.generate(
    model="doubao-seedream-4.0",
    prompt="An orange tabby cat sitting on a windowsill watching the snow, watercolor style",
    size="4096x4096",
    n=1
)

print(response.data[0].url)

Node.js

import OpenAI from 'openai'

const client = new OpenAI({
  baseURL: 'https://open.dieyuyun.com/v1',
  apiKey: 'YOUR_API_KEY',
})

const response = await client.images.generate({
  model: 'doubao-seedream-4.0',
  prompt: 'An orange tabby cat sitting on a windowsill watching the snow, watercolor style',
  size: '4096x4096',
  n: 1,
})

console.log(response.data[0].url)

Response Format

Success Response

{
  "created": 1717000000,
  "model": "doubao-seedream-4.0",
  "data": [
    {
      "url": "https://...",
      "size": "2048x2048"
    }
  ],
  "usage": {
    "generated_images": 1,
    "output_tokens": 16384,
    "total_tokens": 16384
  }
}

Field	Type	Description
created	integer	Unix timestamp (seconds)
model	string	Model ID
data	array	Array of generated results
data[].url	string	Image URL (valid for 24 hours)
data[].b64_json	string	Base64-encoded image
data[].size	string	Actual image dimensions in pixels
usage	object	Usage information
usage.generated_images	integer	Number of images successfully generated
usage.output_tokens	integer	Output token count
usage.total_tokens	integer	Total token count

Error Response

See Error Codes for details.

Best Practices

• Chinese prompts: Doubao Seedream 4.0 has excellent understanding of Chinese prompts, especially in Chinese typography and Eastern aesthetics.
• Style specification: Include explicit style descriptions in your prompt (e.g., "watercolor style", "oil painting", "cyberpunk") for more precise output.
• Multilingual support: Supports both Chinese and English prompts. Keep prompts under 300 Chinese characters for best results.
• Reference images: Pass reference images via the image field to guide the model toward stylistically consistent output. Up to 14 reference images are supported.
• Multi-image mode: Set n greater than 1 or enable sequential_image_generation: "auto" to generate multiple stylistically unified images in a single request.
• Multi-resolution: Use 1K-resolution pixel values for quick previews, 2K for medium-precision scenarios, and 4096x4096 as the default output quality.

Rate Limits

See Rate Limits for details.

Related Documentation

• Doubao Seedream 4.5 - Use the upgraded Doubao Seedream 4.5
• Doubao Seedream 5.0 Lite - Use the lightweight Doubao Seedream 5.0
• Model List & Pricing - Browse available models and pricing
• Playground - Test image generation interactively