Seedream 5.0 API

Use Seedream 5.0 API to generate high-resolution images (up to 3K) with web search integration, custom pixel dimensions, and 11 aspect ratio options for ads, social content, and creative workflows.

Model Type :

Prompt

A modern 5-day weather forecast infographic design for Los Angeles. Horizontal layout. At the top, a clear title reading "Los Angeles 5-Day Weather Forecast". Below the title, there are 5 equal-width vertical cards arranged side-by-side from left to right. Each card clearly displays a daily weather icon (such as sun, clouds), temperature in Fahrenheit (e.g., 75°F), and a clothing suggestion. The clothing suggestions feature young, diverse American men and women in a cartoon illustration style. The characters have natural expressions, relaxed poses, and clear clothing details that match the California weather (e.g., t-shirts, sunglasses, light jackets). The overall style is a modern, clean, and friendly flat vector illustration, UI/UX design aesthetic, clear lines, soft pastel colors, high quality,size 1:1.

Parameters

size

auto

quality

prompt_priority

standard

output_format

jpeg

Estimated Cost (per image)

2.2 Credits

Total = Base Price × Count (n) + Web Search (if enabled)

Sample Result

Prompt

817 (suggested: 2,000)

Input Images

Upload reference images

Click to upload or drag and drop

Supported formats: JPG, JPEG, PNG, WEBP
Maximum file size: 10MB; Maximum files: 10

Size

Quality

Count

For multiple images, describe variations in your prompt

History

Max 20 items

0 running · 0 completed

Your generation history will appear here

Seedream 5.0 API: AI Image Generation with Deep Thinking & Web Search

The first image generation model that thinks before it creates. Seedream 5.0 combines deep reasoning, real-time web search, and up to 14 reference images — available now on EvoLink with competitive pricing.

Billing Rules

•Image generation: base price × number of images (n)
•Web search: charged per request when enabled. A single request may trigger multiple searches.
•All sizes (2K/3K, ratio or custom pixel) share the same image generation price

Pricing

Model	Mode	Price
Seedream 5.0 Lite	Image Generation (All Sizes)	Popular $0.0306/ image(2.2 Credits)
Seedream 5.0 Lite	Web Search (per request)	$0.0069/ request(0.5 Credits)

Popular

Seedream 5.0 Lite

Image Generation (All Sizes)

Price:

$0.0306/ image

(2.2 Credits)

Seedream 5.0 Lite

Web Search (per request)

Price:

$0.0069/ request

(0.5 Credits)

If it's down, we automatically use the next cheapest available—ensuring 99.9% uptime at the best possible price.

What Seedream 5.0 API gives you

Deep Thinking — Reason First, Generate Second

Seedream 5.0 doesn't just follow prompts — it reasons through them. Powered by a unified multimodal architecture that merges understanding, reasoning, and generation into a single model, Deep Thinking breaks down complex instructions before producing a single pixel. Describe a multi-layered infographic with specific data relationships, and the model will parse the logic, resolve ambiguities, and render an image that actually matches your intent.

Try deep thinking generation

Real-Time Web Search — Generate with Live Information

Need an image that references real-world data, current events, or factual content? Seedream 5.0 can search the web during generation to pull in accurate, up-to-date information. Ask for a chart showing the latest market share figures, a poster featuring a real product launch date, or an illustration referencing a recent scientific paper — the model fetches the facts before rendering. This is a standalone capability available at just $0.0069 per request.

Try web search generation

14 Reference Images — Maximum Context, Maximum Control

Seedream 5.0 accepts up to 14 reference images per request, a 40% increase over the 10-image limit in Seedream 4.5. Feed the model your brand assets, style guides, product photos, and layout examples in a single call. The result: outputs that stay on-brand and on-brief without iterative back-and-forth. Ideal for teams that need consistent visual output across campaigns, product lines, or documentation sets.

Try multi-reference generation

Why choose Seedream 5.0 API on evolink.ai

Seedream 5.0 API is exposed through evolink.ai's unified API gateway, allowing teams to integrate image generation using a consistent request and authentication pattern.

Deep Thinking — Reason Before Rendering

Seedream 5.0 reasons through complex prompts before generating. The unified multimodal architecture breaks down multi-element instructions, resolves ambiguities, and produces images that match your intent — not just your keywords.

Real-Time Web Search for Grounded Visuals

Enable web search to let the model fetch current information during generation. Create images that reference real product data, recent events, or factual statistics — ideal for editorial, marketing, and data-driven content.

Competitive Pricing, No Commitments

Seedream 5.0 Lite on EvoLink costs $0.0306 per image vs ~$0.04 through BytePlus. Pay-as-you-go with no minimums. One API integration gives you access to Seedream 5.0 and 40+ other models.

Seedream 5.0 vs 4.5 vs Nano Banana Pro

Feature	Seedream 5.0 Lite	Seedream 4.5	Nano Banana Pro
Price per image	$0.0306	$0.0313	$0.044
Deep Thinking	Yes	No	No
Online Web Search	Yes ($0.0069/req)	No	No
Reference images	Up to 14	Up to 10	Limited
Max resolution	2K / 3K	2K	2K
Scientific / knowledge images	Enhanced	Basic	Basic
Best for	Complex prompts, data-rich visuals	General-purpose	Fast creative

How to integrate Seedream 5.0 API

Seedream 5.0 API follows a simple asynchronous workflow.

Get your API key

Send a generation request

POST to /v1/images/generations with your prompt, model name, size, quality, and optional tools (web_search). You'll receive a task ID immediately.

Poll for results

GET /v1/tasks/{task_id} to check progress. When status is 'completed', download the generated image URLs (valid for 24 hours).

Open Seedream 5.0 API docs

Key features

What makes Seedream 5.0 API stand out

Deep Thinking

Reasons through complex prompts before generating for higher accuracy

Real-Time Web Search

Fetches live information during generation for factually grounded images

14 Reference Images

40% more than Seedream 4.5 for better brand and style consistency

Up to 3K Resolution

2K and 3K outputs for print-ready and high-DPI use cases

11 Aspect Ratios + Custom Pixels

1:1, 9:16, 16:9, 21:9 and more, plus exact pixel control

Unified Multimodal Model

Understanding, reasoning, and generation in a single architecture

Seedream 5.0 API FAQ

Everything you need to know about the product and billing.

Seedream 5.0 is ByteDance's latest AI image generation model with deep thinking and real-time web search capabilities. Unlike traditional image models that directly generate from prompts, Seedream 5.0 reasons about complex instructions before generating, producing more accurate and contextually rich images at up to 3K resolution. It's available through EvoLink's API at $0.0306 per image.

Two major additions: Deep Thinking (the model reasons through complex prompts before generating, improving accuracy for multi-element compositions) and Online Web Search (real-time internet search during generation for factually grounded images). Seedream 5.0 also increases the reference image limit from 10 to 14, adds 3K resolution support, offers 11 aspect ratio presets plus custom pixel dimensions, and improves scientific/knowledge image generation.

Seedream 5.0 Lite costs $0.0306 per image (2.2 EvoLink Credits). The Web Search add-on is $0.0069 per request (0.5 Credits). BytePlus official pricing is approximately $0.04 per image, so you save roughly 23% through EvoLink. All pricing is pay-as-you-go with no minimum commitments.

Yes, and it's significantly better at it than previous Seedream versions. The Deep Thinking capability allows Seedream 5.0 to understand structural relationships in scientific content — molecular bonds, process flows, data hierarchies — and render them with logical coherence. Combine this with Web Search to ground your scientific visuals in real, current data.

When you enable web_search: true in your API request, Seedream 5.0 searches the web for relevant information related to your prompt before generating the image. The model can incorporate real product details, current statistics, recent events, or factual data into the generated visual. The web search capability costs $0.0069 per request on EvoLink.

Yes. Seedream 5.0 supports up to 14 reference images per API request — a 40% increase over Seedream 4.5's 10-image limit. You can pass brand assets, style references, product photos, layout examples, or any combination to guide the model's output. This is especially useful for maintaining visual consistency across campaigns or generating variations that stay on-brand.

Explore more image APIs on EvoLink

Nano Banana 2|Nano Banana Pro|Seedream 4.0|Seedream 4.5

API Reference

Select endpoint

Authentication

All APIs require Bearer Token authentication.

Header

Authorization: 
Bearer YOUR_API_KEY

Get API Key

POST

/v1/images/generations

Generate Image

Create an image generation task using text prompts with optional web search enhancement. Supports ratio-based and custom pixel size outputs in 2K/3K quality.

Asynchronous processing mode, use the returned task ID to .

Generated image links are valid for 24 hours, please save them promptly.

Request Parameters

modelstringRequiredDefault: doubao-seedream-5.0-lite

Image generation model name.

Value	Description
doubao-seedream-5.0-lite	Seedream 5.0 Lite model

Exampledoubao-seedream-5.0-lite

promptstringRequired

Prompt describing the image to be generated.

Notes

Limited to 2000 tokens

ExampleA serene lake reflecting the beautiful sunset

nintegerOptionalDefault: 1

Maximum number of images to generate.

Value	Description
1-15	Any integer between 1 and 15

Notes

Each request will pre-charge based on the value of n, actual charges based on the number of images generated

Example1

sizestringOptionalDefault: 2048x2048

Size of generated images. Supports aspect ratios or custom pixel dimensions.

Value	Description
auto	Auto - model determines optimal ratio
1:1	Square
2:3 / 3:2	Portrait / Landscape
3:4 / 4:3	Portrait / Landscape
4:5 / 5:4	Portrait / Landscape
9:16 / 16:9	Vertical / Horizontal
21:9	Ultra-wide
WIDTHxHEIGHT	Custom pixel dimensions (e.g. 1920x1080)

Notes

Custom pixel dimensions must be within the valid range for the selected quality tier
Aspect ratio mode: backend auto-converts to optimal pixel dimensions based on quality setting

Example16:9

qualitystringOptionalDefault: 2K

Output quality tier, affects resolution of generated images.

Value	Description
2K	High quality (default)
3K	Ultra quality

Example2K

prompt_prioritystringOptionalDefault: standard

Prompt optimization strategy.

Value	Description
standard	Default priority, higher quality

Examplestandard

image_urlsarrayOptional

Reference image URL list for image-to-image and image editing.

Notes

Max 14 images per request
Max file size: 10MB per image
Supported formats: .jpeg, .jpg, .png, .webp, .bmp, .tiff, .gif
Aspect ratio (width/height): [1/16, 16]
Max total pixels: 6000×6000

Example["https://example.com/image1.png"]

model_params.output_formatstringOptionalDefault: jpeg

Output image format. Nested inside model_params object.

Value	Description
jpeg	Compressed format, smaller file size (default)
png	Lossless format

Examplejpeg

model_params.toolsarrayOptional

Optional tools to enhance generation. Nested inside model_params object.

Notes

Web search enables the model to reference online content for more accurate results
Web search incurs an additional per-request charge (SKU 1330)
Format: {"model_params": {"tools": [{"type": "web_search"}]}}

Example[{"type": "web_search"}]

Examples

Request Example

{
  "model": "doubao-seedream-5.0-lite",
  "prompt": "A futuristic cityscape at sunset with flying cars",
  "size": "16:9",
  "quality": "2K",
  "n": 1,
  "prompt_priority": "standard",
  "model_params": {
    "output_format": "jpeg",
    "tools": [{"type": "web_search"}]
  }
}

Response Example

{
  "created": 1772000000,
  "id": "task-unified-1772000000-seedream5lite",
  "model": "doubao-seedream-5.0-lite",
  "object": "image.generation.task",
  "progress": 0,
  "status": "pending",
  "task_info": {
    "can_cancel": true,
    "estimated_time": 45
  },
  "type": "image",
  "usage": {
    "billing_rule": "per_call",
    "credits_reserved": 2.2,
    "user_group": "default"
  }
}