ByteDance Seedream Image Generation
The strongest image generation model under ByteDance.
Your API key is pre-filled in the snippet — copy and run it in your terminal or IDE.
This is only a basic call example. See the full docs for more parameters and advanced usage.
Seedream API
ByteDance AI Image Generation
Generate images using ByteDance Seedream model through simple API calls. Supports text-to-image, image editing—up to 4K resolution, multiple model options, pay-as-you-go. ```
Why Choose Seedream?
Seedream is an AI image generation model launched by ByteDance, known for its excellent image quality, precise text understanding, and rich aesthetic expression. Seedream 4.5 ranks among the industry leaders in multiple image generation benchmarks, particularly excelling in understanding Chinese prompts.
Ace Data Cloud provides comprehensive Seedream API proxy services—you do not need to apply for a ByteDance API account or resolve regional restrictions. High-quality images can be generated through a simple REST API, supporting both text generation and image editing capabilities.
Core Features of Seedream API
From text descriptions to exquisite images—Seedream offers comprehensive image generation and editing capabilities
Text-to-Image (T2I)
Generate high-quality images by inputting text descriptions. Supports both Chinese and English prompts, with precise understanding of complex semantics and detailed, realistic visuals.
Image Editing (I2I)
Upload the original image and describe the editing intent; the SeedEdit model can intelligently modify the image content while maintaining overall style consistency.
Multi-Resolution Support
Supports 1K, 2K, 4K, and adaptive multiple resolutions. Flexibly meets different needs from quick previews to print-quality standards.
Sequential Image Generation
Enable the sequential_image_generation option to generate a series of stylistically consistent images, suitable for comics and storyboards.
Streaming Output
Supports stream mode, returning generation results step by step. Significantly optimizes user waiting experience in high-resolution scenarios.
Watermark Control
Flexibly control whether to add AI-generated watermarks through the watermark parameter, meeting compliance needs for different usage scenarios.
import requests
response = requests.post(
"https://api.acedata.cloud/seedream/images",
headers={
"Authorization": "Bearer YOUR_API_KEY",
"Content-Type": "application/json"
},
json={
"model": "doubao-seedream-4-5-251128",
"prompt": "a white siamese cat"
}
)
data = response.json()
print(data)
{
"success": true,
"task_id": "c0aae74d-4442-4563-b620-3da86bae5434",
"trace_id": "581c4721-83e2-41a3-b7d8-b64378604bed",
"data": [
{
"prompt": "a white siamese cat",
"size": "2048x2048",
"image_url": "https://platform.cdn.acedata.cloud/seedream/819386b9-fab6-4bf3-acc6-df29f53c7248.jpg"
}
]
}
</div>
</div>
<div class="code-right">
<h2>Generate AI Images with One Line of Code</h2>
<p>Send a POST request with the text prompt to instantly obtain the generated high-quality image.</p>
<div class="explain-steps">
<div class="explain-step">
<div class="step-num">1</div>
<div class="step-text">
<h4>Prepare Request Parameters</h4>
<p>Select the model and provide the text prompt <code>prompt</code>, optional resolution and seed parameters</p>
</div>
</div>
<div class="explain-step">
<div class="step-num">2</div>
<div class="step-text">
<h4>Send Generation Request</h4>
<p>POST to <code>/seedream/images</code>, directly returning the image URL of the generated result</p>
</div>
</div>
<div class="explain-step">
<div class="step-num">3</div>
<div class="step-text">
<h4>Or Use Callback Notification</h4>
<p>Set the <code>callback_url</code> to actively POST to your server after generation is complete</p>
</div>
</div>
</div>
</div>
</div>
What can Seedream be used for?
From creative design to commercial applications—Seedream is being used in these scenarios
Social Media Content
Quickly generate eye-catching social media images, cover photos, and post illustrations
E-commerce Materials
Batch generate product display images, scene images, and advertising creative materials, reducing design costs
Game & Concept Design
Generate game characters and scene concept art—quickly validate creative and visual style directions
Content Illustrations
Automatically generate matching high-quality illustrations for articles, blogs, news, and other content
Quick Start in 3 Steps
From registration to generating the first image, it only takes a few minutes
Register and Get API Key
Create a free account on Ace Data Cloud and generate your Bearer Token from the console.
Send Image Generation Request
POST to /seedream/images, passing in the text prompt and model name.
Get Generation Results
Directly obtain the generated image URL from the response, download and use it.
Comparison with Direct Access to ByteDance API
Ace Data Cloud allows you to focus on usage without worrying about registration, authentication, and regional restrictions
| Comparison Item | Ace Data Cloud | ByteDance Direct Connection |
|---|---|---|
| Registration Threshold | Free registration, immediate use | Need to apply for a ByteDance account |
| Regional Restrictions | Available globally | ✗ Some regions are restricted |
| API Style | Unified REST API | ByteDance native SDK |
| Billing Method | Pay-as-you-go, recharge to use | Need to sign a contract to activate |
| Technical Support | 7×24 online support | Ticket-based |
Select the Right Seedream Model
From text-to-image to image editing—multiple models cover different creative needs
Seedream 4.5
Latest FlagshipThe latest flagship model of Seedream. Comprehensive upgrades in image quality, text understanding, and aesthetic expression. Supports continuous image generation and streaming output.
- ✓ Best image quality
- ✓ 1K / 2K / 4K / Adaptive
- ✓ Continuous image generation
- ✓ Streaming output
- ✓ $0.037 / image
Seedream 4.0
Stable VersionA mature and stable main model. Achieves the best balance between image quality and cost-effectiveness, suitable for most production scenarios.
- ✓ High-quality images
- ✓ 1K / 2K / 4K / Adaptive
- ✓ Continuous image generation
- ✓ Streaming output
- ✓ $0.030 / image
Seedream 3.0 T2I
Text-to-ImageThe third generation text-to-image model. Supports precise control of seed values and guidance coefficients, suitable for batch generation scenarios that require reproducibility.
- ✓ Text generates images
- ✓ Seed controllable reproducibility
- ✓ Guidance coefficient adjustment
- ✓ $0.038 / image
SeedEdit 3.0 I2I
Image EditingDedicated image editing model. Upload the original image and describe the editing intent with text to intelligently modify the local or overall style of the image.
- ✓ Intelligent image editing
- ✓ Seed controllable reproducibility
- ✓ Guidance coefficient adjustment
- ✓ $0.046 / image
Seedream API Pricing
Billing is based on the number of images generated, with prices depending on the model. The task query interface is completely free.
| Model | Type | Unit Price (per image) |
|---|---|---|
| Seedream 4.5 | Text-to-Image | $0.037 |
| Seedream 4.0 | Text-to-Image | $0.030 |
| Seedream 3.0 T2I | Text-to-Image | $0.038 |
| SeedEdit 3.0 I2I | Image Editing | $0.046 |
The task query interface /seedream/tasks is completely free
Frequently Asked Questions
Everything you need to know about using the Seedream API
What is Seedream? ▾
Seedream is a series of AI image generation models launched by ByteDance, including text-to-image (T2I) and image editing (I2I) capabilities. This model is known for its excellent text understanding, visual aesthetics, and detail representation, supporting output up to 4K resolution.
Does it support image editing? ▾
Yes! Using the doubao-seededit-3-0-i2i-250628 model, pass the URL or Base64 encoding of the original image in the image parameter, and describe the editing intent through prompt to achieve intelligent image editing. Supports local modifications and overall style transformations.
How long does image generation take? ▾
In most cases, Seedream can complete image generation within 5-15 seconds. The specific time depends on the model version and output resolution. 4K resolution images will take longer than 1K resolution. You can also use callback notifications (callback_url) to avoid blocking waits.
What resolutions are supported? ▾
Seedream 4.5 and 4.0 support 1K, 2K, 4K, and adaptive resolutions. You can also specify exact pixel dimensions (e.g., 1024x1024). The default output is 2048x2048.
How to choose a model? ▾
It is recommended to use Seedream 4.5 for the best image quality. Seedream 4.0 is the cost-effective choice ($0.030/image). If you need image editing functionality, use SeedEdit 3.0. Seedream 3.0 T2I is suitable for batch generation scenarios that require seed value reproducibility.
What is continuous image generation? ▾
By setting sequential_image_generation: "auto", Seedream 4.5 and 4.0 models will generate a series of images with consistent styles and related content based on your input. Suitable for scenarios requiring visual consistency, such as comic storyboards, storyboards, and product series displays.
Can the watermark be turned off? ▾
Yes. Set watermark: false in the request to turn off the AI-generated watermark. By default, the watermark is enabled. Please use AI-generated content identifiers reasonably according to local regulations.
Explore More AI Image and Video Services
We offer a variety of AI creative tools—from images to videos to music
Seedance API
ByteDance AI video generation—text-to-video, image-to-video, same series of products
Midjourney API
World-class AI image generation—create beautiful images through text prompts
Flux API
Black Forest Labs image generation—high-speed, high-quality AI drawing
Suno API
AI music generation—create custom songs and pure music through API
Start Using Seedream API Now
Generate high-quality AI images through simple API calls. Pay-as-you-go—no subscription fees, no commitments.
ByteDance Seedream image generation API allows you to input text to generate images.
ByteDance Seedream Task Query API
Packages
We offer various credit packages. The more you purchase, the higher the discount you receive.