This MCP server enables AI image generation, editing, and decomposition with ByteDance Seedream models, plus task tracking and model/size info.
Generate new images from text prompts (Chinese/English) using v5.0 Pro/Lite, v4.5, or v4.0 models.
Edit existing images: style transfer, background changes, attribute edits, virtual try-on, and transparent-background Pro edits.
Decompose an image into a base image and up to 16 editable transparent PNG layers.
Use advanced options: multi-resolution sizes, sequential generation, web search, prompt optimization, watermark, output/response formats, and callbacks.
List available models with capabilities/pricing and list available image sizes.
Query single or multiple asynchronous tasks to retrieve status, image URLs, and metadata.
Provides access to ByteDance's Seedream models for AI image generation and editing, including text-to-image creation, style transfer, background modification, and virtual try-on capabilities.
A Model Context Protocol (MCP) server for AI image generation and editing using ByteDance's Seedream models through the AceDataCloud API.
Generate and edit AI images directly from Claude, VS Code, or any MCP-compatible client.
Text-to-Image Generation — Create high-quality images from text prompts (Chinese & English)
Image Editing — Modify existing images, including transparent-background Pro edits
Multiple Models — Seedream v5.0 (flagship), v4.5, and v4.0
Layer Decomposition — Split one image into a base plus up to 16 editable transparent PNG layers
Multi-Resolution — Pro 1K/1.5K/2K; Lite 2K/3K/4K; custom dimensions
Sequential Generation — Generate related images in sequence (5.0 Lite/4.5/4.0)
Web Search — Let 5.0 Lite use current web information when needed
Task Tracking — Monitor generation progress and retrieve results
Related MCP server: Doubao Image/Video Generation MCP Server
Tool | Description |
| Generate an AI image from a text prompt using ByteDance's Seedream model. |
| Edit images, including Seedream 5.0 Pro transparent-background mode. |
| Split one image into a base and up to 16 positioned transparent layers. |
| Query the status and result of a Seedream image generation or edit task. |
| Query multiple Seedream image tasks at once. |
| List all available Seedream models with their capabilities and pricing. |
| List all available image sizes and resolution options for Seedream. |
Capability | 5.0 Pro | 5.0 Lite |
Single image generation/editing | Yes | Yes |
Layer decomposition / transparent background | Yes | No |
Sequential images / web search | No | Yes |
Prompt optimization | standard, fast | standard |
Preset sizes | 1K, 1.5K, 2K | 2K, 3K, 4K |
MCP image tools submit asynchronously and return a task id. Use seedream_get_task until completion. For real-time NDJSON streaming, call the REST API or CLI rather than combining stream with MCP async submission.
The hosted endpoint is https://seedream.mcp.acedata.cloud/mcp. Choose one route for the MCP client:
Route | When to use it | Credential setup |
Hosted OAuth | The client supports remote MCP OAuth | Add only the URL, then sign in to AceDataCloud and approve access. No token needs to be pasted into client configuration. |
Hosted API token | The client cannot finish OAuth, or you need an explicit integration credential | Send an AceDataCloud API token in the |
Local stdio | The client runs a local MCP process | Install |
The hosted service advertises OAuth metadata and Dynamic Client Registration (DCR). DCR registers the client application; it is not an API key. OAuth signs you in and the client sends the resulting Bearer token; it may reuse or create an API credential for the account. Browser sign-in still requires an AceDataCloud account. The hosted service can be metered: review current service documentation and displayed pricing before a real operation. Do not configure both an OAuth login and a fixed Authorization header for the same server.
Claude and Claude Desktop chat: Add a remote custom connector in Customize → Connectors → Add custom connector, enter https://seedream.mcp.acedata.cloud/mcp, select sign-in, and choose Register automatically if Claude asks how to register its OAuth client. Complete consent. Claude Desktop's local claude_desktop_config.json is a separate setup. Claude connector guide.
Claude Code: claude mcp add --transport http --scope user seedream https://seedream.mcp.acedata.cloud/mcp, then claude mcp login seedream. Check /mcp. Claude Code MCP guide.
Cursor: Add a remote server with only https://seedream.mcp.acedata.cloud/mcp. For a project, merge the entry below into <project>/.cursor/mcp.json; for personal use, use ~/.cursor/mcp.json. Cursor MCP guide.
VS Code / Copilot: Run MCP: Add Server, select HTTP, enter https://seedream.mcp.acedata.cloud/mcp, then finish the browser sign-in. New portable workspace configs use <project>/.mcp.json; the VS Code-specific format below uses <project>/.vscode/mcp.json or the user profile. Check MCP: List Servers. VS Code MCP setup.
Codex: codex mcp add seedream --url https://seedream.mcp.acedata.cloud/mcp, then codex mcp login seedream. Its user settings are in ~/.codex/config.toml. Official Codex MCP guide.
Cursor project config (OAuth):
{
"mcpServers": {
"seedream": {"url": "https://seedream.mcp.acedata.cloud/mcp"}
}
}VS Code-specific workspace config (OAuth):
{
"servers": {
"seedream": {"type": "http", "url": "https://seedream.mcp.acedata.cloud/mcp"}
}
}Open Ace Data Cloud → Applications. Sign in, or register an account first.
Find General application. If it has not been created, choose Initialize General Application. A general key can use the services available to your account; a dedicated application's key is limited to its service. Check the required service/model access and current pricing.
Copy the API Key from the application card, or open Keys / Manage Keys → Create to create a separate key for this integration. Give it a recognizable name so you can revoke or rotate it independently.
When creating a key, review Expiration Time, Set Usage Limitation, and Allowed APIs. Enable only the restrictions you need, allow the operations you will call, then click Create and copy the token from the key list. Task lookup does not grant permission to submit a generation. Confirm the application's balance can cover your first request; any trial allowance is whatever your account currently displays.
Copy only the token value, without quotation marks, spaces, or Bearer . An API Key / API Token field expects this raw value; only a manually configured HTTP Authorization header needs the Bearer prefix. A platform-... account-management token cannot replace a generation API key.
Save the key in the client's credential store or the environment variable shown below. Never put it in a prompt, shared workflow, screenshot, or Git commit. If authentication fails, check the full copied value, application/service access, expiration, usage limit, and balance before creating another key.
A fixed Bearer header bypasses OAuth. If the header is invalid, remove it before trying browser sign-in; keep the saved credential private.
For Claude Code, the shell expands the token when you add the server; treat the saved user MCP config as a secret:
export ACEDATACLOUD_API_TOKEN='YOUR_API_TOKEN'
claude mcp add --transport http --scope user seedream https://seedream.mcp.acedata.cloud/mcp \
--header "Authorization: Bearer $ACEDATACLOUD_API_TOKEN"For a Claude Code project config, put a variable reference in <project>/.mcp.json and set that variable in the environment that launches Claude Code:
{
"mcpServers": {
"seedream": {
"type": "http",
"url": "https://seedream.mcp.acedata.cloud/mcp",
"headers": {"Authorization": "Bearer ${ACEDATACLOUD_API_TOKEN}"}
}
}
}Cursor uses a different environment-variable syntax in ~/.cursor/mcp.json or an uncommitted project config:
{
"mcpServers": {
"seedream": {
"url": "https://seedream.mcp.acedata.cloud/mcp",
"headers": {"Authorization": "Bearer ${env:ACEDATACLOUD_API_TOKEN}"}
}
}
}In VS Code, run MCP: Open User Configuration and merge this server plus its masked input; ${input:...} is for VS Code's user/workspace format and is not portable to the Agent Host .mcp.json format:
{
"inputs": [
{"id": "acedata-seedream-token", "type": "promptString", "description": "AceDataCloud API token", "password": true}
],
"servers": {
"seedream": {
"type": "http",
"url": "https://seedream.mcp.acedata.cloud/mcp",
"headers": {"Authorization": "Bearer ${input:acedata-seedream-token}"}
}
}
}For Cline, use its MCP configuration UI or CLI file ~/.cline/data/settings/cline_mcp_settings.json; its remote transport value is streamableHttp. For JetBrains AI Assistant, add a remote URL from Settings → Tools → AI Assistant → Model Context Protocol (MCP). For Zed, use a context_servers entry with the URL only for OAuth or add a local Bearer header. These clients have different configuration schemas; follow their current UI rather than copying another client's JSON. Cline · JetBrains · Zed.
Install the package and give the local process an API token:
python -m pip install mcp-seedream-pro
export ACEDATACLOUD_API_TOKEN='YOUR_API_TOKEN'
mcp-seedream-proFor Claude Desktop local MCP, merge this entry into the file opened by its developer settings (~/Library/Application Support/Claude/claude_desktop_config.json on macOS). uvx requires uv on PATH:
{
"mcpServers": {
"seedream": {
"command": "uvx",
"args": ["mcp-seedream-pro"],
"env": {"ACEDATACLOUD_API_TOKEN": "YOUR_API_TOKEN"}
}
}
}Keep this user-level file private. Self-hosted HTTP uses mcp-seedream-pro --transport http --port 8000; expose it only with suitable network and TLS controls. Local execution still calls the AceDataCloud API.
https://seedream.mcp.acedata.cloud/health returning {"status":"ok"} checks endpoint reachability only.
Confirm that the MCP client loads tools. seedream_list_models is a reference tool; it does not verify downstream API access or balance.
If you need a full API check, call seedream_generate_image with your own valid input after reviewing current service documentation and displayed pricing. If the result contains a task ID, call seedream_get_task on that same ID until terminal success or failure. Do not resubmit the operation just to check progress.
For 401, check which auth route the client used and whether the token or OAuth session is valid. A 403 may mean an account permission or content moderation failure; read the returned error. Insufficient balance and downstream service failures need their own diagnosis. A listed tool or submitted task does not prove a successful result.
Tool | Description |
| Generate an image from a text prompt |
| Edit or modify existing images with AI |
Tool | Description |
| Query a single task status and result |
| Query multiple tasks at once |
Tool | Description |
| List available models with capabilities |
| List available image size options |
Model | Version | Type | Best For | Price |
| v5.0 Pro | Generate/Edit | Single image, transparent background, layer decomposition | Tiered Credits |
| v5.0 Lite | Text-to-Image | Sequential generation, streaming, web search | See live pricing |
| v4.5 | Text-to-Image | Previous flagship, great quality | See live pricing |
| v4.0 | Text-to-Image | Best value, most tasks | See live pricing |
User: Create a photorealistic image of a cat in a garden
Claude: I'll generate that image for you.
[Calls seedream_generate_image with detailed prompt]
→ Returns task_id and image URLUser: Change the background of this photo to a beach
[Provides image URL]
Claude: I'll edit that image for you.
[Calls seedream_edit_image with image URL and edit description]User: 生成一幅中国山水画，有远山、流水和古松
Claude: 好的，我来为您生成这幅山水画。
[Calls seedream_generate_image with Chinese prompt]Variable | Description | Default |
| API token from AceDataCloud | Required |
| API base URL |
|
| OAuth client ID (hosted mode) | — |
| Platform base URL |
|
| Request timeout in seconds |
|
| Logging level |
|
mcp-seedream-pro --help
Options:
--version Show version
--transport Transport mode: stdio (default) or http
--port Port for HTTP transport (default: 8000)# Clone repository
git clone https://github.com/AceDataCloud/SeedreamMCP.git
cd SeedreamMCP
# Create virtual environment
python -m venv .venv
source .venv/bin/activate # or `.venv\Scripts\activate` on Windows
# Install with dev dependencies
pip install -e ".[dev,test]"# Run unit tests
pytest
# Run with coverage
pytest --cov=core --cov=tools
# Run integration tests (requires API token)
pytest -m integration# Format code
ruff format .
# Lint code
ruff check .
# Type check
mypy core tools main.py# Install build dependencies
pip install -e ".[release]"
# Build package
python -m build
# Upload to PyPI
twine upload dist/*SeedreamMCP/
├── core/ # Core modules
│ ├── __init__.py
│ ├── client.py # HTTP client for Seedream API
│ ├── config.py # Configuration management
│ ├── exceptions.py # Custom exceptions
│ ├── server.py # MCP server initialization
│ ├── types.py # Type definitions
│ └── utils.py # Utility functions
├── tools/ # MCP tool definitions
│ ├── __init__.py
│ ├── image_tools.py # Image generation/editing tools
│ ├── task_tools.py # Task query tools
│ └── info_tools.py # Model & size info tools
├── prompts/ # MCP prompt templates
│ └── __init__.py
├── tests/ # Test suite
│ ├── conftest.py
│ ├── test_config.py
│ └── test_utils.py
├── deploy/ # Deployment configs
│ ├── run.sh
│ └── production/
│ ├── deployment.yaml
│ ├── ingress.yaml
│ └── service.yaml
├── .github/ # GitHub Actions workflows
│ ├── dependabot.yml
│ └── workflows/
│ ├── ci.yaml
│ ├── claude.yml
│ ├── deploy.yaml
│ └── publish.yml
├── .env.example # Environment template
├── .gitignore
├── .ruff.toml # Ruff linter config
├── CHANGELOG.md
├── Dockerfile # Docker image for HTTP mode
├── docker-compose.yaml # Docker Compose config
├── LICENSE
├── main.py # Entry point
├── pyproject.toml # Project configuration
└── README.mdThis server wraps the AceDataCloud Seedream API:
Seedream Images API — Image generation and editing
Seedream Tasks API — Task queries
AI Art Creation — Generate stunning artwork, illustrations, and digital art
Product Photography — Create professional product scene compositions
Content Creation — Generate images for blogs, social media, marketing
Virtual Try-On — Visualize clothing on different models
Style Transfer — Transform photos into different art styles
Game Design — Concept art, character design, environment design
E-commerce — Product mockups, lifestyle shots, banner images
MIT License - see the LICENSE file for details.
Decompose one image into a base image and up to 16 editable transparent layers.
| Name | Required | Description | Default |
|---|---|---|---|
| size | No | Output size: auto, 1K, 1.5K, or 2K. | auto |
| image | Yes | One PNG or JPEG URL/base64 image to decompose. | |
| prompt | No | Optional elements to decompose; omit for automatic decomposition. Supports <bbox> coordinates. | |
| watermark | No | Whether to add an AI-generated watermark. | |
| callback_url | No | Optional webhook URL for async delivery. | |
| output_format | No | Base image format; layers are always PNG. | jpeg |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden. It usefully discloses the output structure (base image plus up to 16 transparent layers), but it does not disclose side effects or caveats such as async task behavior, default watermarking, non-destructiveness, or cost/rate implications.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
One sentence contains a verb, object, and output shape with no filler. The key information is front-loaded, and every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a six-parameter tool with an output schema and no annotations, the description is serviceable but incomplete: it omits the asynchronous/deferred result path implied by callback_url and the need to poll via seedream_get_task, and it does not address watermark or task lifecycle behavior. However, the input schema is thorough and the output schema is present, closing some gaps.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the input schema already documents all six parameters and the enum. The top-level description adds no parameter-level meaning beyond the schema, which matches the baseline of 3.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Decompose') with a clear resource and outcome: one input image becomes a base image plus up to 16 editable transparent layers. This distinguishes it from sibling tools like seedream_edit_image or seedream_generate_image even without explicitly naming them.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The intended use is implied by the verb and the layer output, but the description does not explicitly state when an agent should choose decompose over edit_image, nor does it mention exclusions or prerequisites. Sibling tool names provide some context, but the description itself gives no direct routing guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Edit or modify existing images using ByteDance's Seedream/SeedEdit model.
This tool modifies existing images based on text instructions. It can change
styles, backgrounds, attributes, clothing, and more. Supports single or
multiple image inputs.
Use this when:
- You want to modify or transform an existing image
- You need to change style, background, colors, or attributes
- You want to apply artistic transformations (watercolor, oil painting, etc.)
- You need virtual try-on (clothing on person)
- You want to place objects in different scenes
Common use cases:
- Style transfer: "Convert to anime style", "Make it look like a pencil sketch"
- Background change: "Replace background with a sunset beach"
- Attribute edit: "Change hair color to blonde", "Add sunglasses"
- Virtual try-on: Provide person image + clothing image
- Scene composition: Place products in realistic environments
Returns:
JSON with task_id, trace_id, success status, and edited image data
including image URLs.
| Name | Required | Description | Default |
|---|---|---|---|
| size | No | Model-specific output size or explicit dimensions. Pro supports 1K/1.5K/2K; Lite supports 2K/3K/4K. | |
| image | Yes | List of image URLs or base64-encoded images to edit. Supports HTTP/HTTPS URLs (publicly accessible) or base64 format (data:image/png;base64,...). Each image must be under 10MB. | |
| model | No | Model to use for editing. Seedream 5.0 Pro, 5.0 Lite, 4.5, and 4.0 all support image editing when images are provided. | doubao-seedream-5-0-lite-260128 |
| tools | No | Optional list of tool types for the model to use during editing. | |
| prompt | Yes | Description of the edit to perform on the image(s). Describe what changes you want. Example: 'Change the background to a beach scene', 'Make the person wear a red dress', 'Convert to watercolor painting style' | |
| watermark | No | Whether to add an AI-generated watermark. Default is true. | |
| background | No | Seedream 5.0 Pro background mode. transparent requires one PNG input and PNG output. | |
| callback_url | No | Optional webhook URL for async result notification. | |
| output_format | No | Output image format. 'jpeg' (default) or 'png'. | |
| response_format | No | Response format. 'url' (default) or 'b64_json'. | |
| optimize_prompt_options | No | Prompt optimization. Pro supports standard/fast; Lite supports standard. | |
| sequential_image_generation | No | Generate related images based on input. 'auto' enables it. | |
| sequential_image_generation_options | No | Tunable options for grouped image generation. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden, and it does disclose the return payload (task_id, trace_id, success, image URLs), watermark defaulting to true, and the PNG constraint for transparent background. However it never says whether the call is synchronous or asynchronous (the mere existence of callback_url implies async), nor mentions permissions, cost, latency, or failure modes for a 13-parameter generation job.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Purpose and scope are front-loaded, which is good, but the 'Use this when' bullets and the 'Common use cases' bullets substantially restate the same ground (style, background, attributes, virtual try-on), so several sentences do not fully earn their place. Structure is readable but repetitive.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With 13 parameters, an output schema already present, and 100% schema coverage, the description supplies what structured fields cannot: the intent-level taxonomy of edits and multi-image input semantics. The main remaining gap is async/execution behavior, but for a tool of this complexity the description is largely complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents size, model, background, output_format and the rest; the description only adds illustrative prompt examples ('Make the person wear a red dress'). Per the coverage rule this is a baseline 3, and the description does not add much beyond the schema for the remaining parameters.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb and resource ('Edit or modify existing images') and immediately scopes it to image-to-image transformation driven by text instructions. The framing as modification of an existing image implicitly but clearly separates it from the sibling seedream_generate_image, and an agent can tell which one to pick without opening a schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
A dedicated 'Use this when' block plus five concrete use-case bullets give strong positive guidance (style transfer, background change, attribute edit, virtual try-on, scene composition). It never names an alternative tool or states a when-not condition (e.g. use generate_image when no source image exists), so it stops short of 5.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Generate an AI image from a text prompt using ByteDance's Seedream model.
This tool creates high-quality images from text descriptions using ByteDance's
Seedream models (powered by Doubao). Supports multiple model versions with different
capabilities and quality levels.
Use this when:
- You want to generate a new image from scratch based on a text description
- You need high-quality AI-generated images (photos, illustrations, art)
- You want to create images with specific styles, compositions, or themes
Do NOT use this when:
- You want to edit or modify an existing image (use seedream_edit_image instead)
- You need to combine multiple images (use seedream_edit_image instead)
Model selection guide:
- v5.0 Lite (doubao-seedream-5-0-lite-260128): Image sets, streaming, and web search
- v4.5 (doubao-seedream-4-5-251128): Previous flagship, great quality and detail
- v4.0 (doubao-seedream-4-0-250828): Stable and cost-effective, great for most tasks
Returns:
JSON with task_id, trace_id, success status, and generated image data
including image URLs.
| Name | Required | Description | Default |
|---|---|---|---|
| size | No | Model-specific output size: Pro supports 1K/1.5K/2K (and auto for decomposition); Lite supports 2K/3K/4K. Explicit dimensions such as 2048x1024 are also accepted. | |
| model | No | Model to use for generation. 'doubao-seedream-5-0-pro-260628' (v5.0 Pro, flagship single image, highest quality; no sequential generation, streaming, or web search). 'doubao-seedream-5-0-lite-260128' (v5.0 Lite, sequential generation, streaming, web search). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value). | doubao-seedream-5-0-lite-260128 |
| tools | No | Optional list of tool types for the model to use during generation. Currently only 'web_search' is supported. Only supported by doubao-seedream-5-0-lite-260128 (v5.0 Lite). | |
| prompt | Yes | Description of the image to generate. Be descriptive about subject, style, atmosphere, lighting, and composition. Supports both Chinese and English. Example: 'A photorealistic portrait of an astronaut on Mars, golden hour lighting, cinematic composition, ultra-detailed' | |
| watermark | No | Whether to add an AI-generated watermark. Default is true. | |
| callback_url | No | Optional webhook URL to receive the result asynchronously. The API will POST the result to this URL when complete. Must be publicly accessible. | |
| output_format | No | Output image format. 'jpeg' (default) or 'png'. | |
| response_format | No | Response format for the generated image. 'url' (default) returns a public image URL. 'b64_json' returns base64-encoded image data. | |
| optimize_prompt_options | No | Optional prompt optimization configuration. Supports `mode` with values 'standard' (higher quality, slower) or 'fast' (quicker, lower quality). Only supported on doubao-seedream-4.5 (standard mode only) and doubao-seedream-4.0. | |
| sequential_image_generation | No | Generate related images based on input. 'auto' enables it, 'disabled' (default) turns it off. Supported by Seedream 5.0 Lite, 4.5, and 4.0; not Pro. | |
| sequential_image_generation_options | No | Tunable options for grouped image generation. Only honored when `sequential_image_generation=auto`. Supports `max_images` (int, range [1, 15]). Supported by Seedream 5.0 Lite, 4.5, and 4.0; not Pro. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden, and it does disclose the async task model (task_id/trace_id, callback_url), the model capability differences (streaming, sequential generation, web search only on Lite/4.5/4.0, not Pro), and the return payload shape. It stops short of stating auth requirements, rate limits, or generation latency, so a small gap remains.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loaded with the core action, then organized into scannable when/when-not/model blocks and a Returns section. It is slightly long and the model guide duplicates schema enum text, but nothing is unearned.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For an 11-parameter generation tool with a full output schema, the description covers the generation intent, model trade-offs, and async return shape adequately. It omits practical operational context such as authentication, billing, or typical generation time, which an agent may still need.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the baseline is 3. The model selection guide partially restates schema enum descriptions (e.g. Lite supports streaming/web search) rather than adding net-new semantics, though it helps consolidate capability trade-offs in one place.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb (generate) and resource (AI image) plus the underlying model family, and explicitly distinguishes itself from seedream_edit_image for the editing/combining cases. An agent can identify this as the text-to-image entry point without opening another schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides explicit 'Use this when' and 'Do NOT use this when' blocks, naming seedream_edit_image as the alternative for modification and multi-image combination. The model selection guide further routes among capability tiers.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Query the status and result of a Seedream image generation or edit task.
Use this to check if a generation/edit is complete and retrieve the resulting
image URLs and metadata.
Use this when:
- You want to check if an image generation has completed
- You need to retrieve image URLs from a previous generation
- You want to get the full details of a generated/edited image
Returns:
Task status and image information including URLs, prompts, and metadata.
| Name | Required | Description | Default |
|---|---|---|---|
| task_id | Yes | The task ID returned from a generation or edit request. This is the 'task_id' field from any seedream_generate_image or seedream_edit_image tool response. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Discloses it returns status and image information, implying a read-only operation. With no annotations, the description carries the burden and does well, though it could explicitly state it is non-destructive.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Concisely structured with bulleted usage scenarios and a return summary, though some minor repetition between the first line and the list.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present and a single required parameter, the description fully covers the tool's purpose and usage, leaving no gaps.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% and the description's parameter info duplicates the schema description, adding no new meaning beyond what the schema provides.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it queries the status and result of a specific task, distinguishing it from sibling tools like seedream_generate_image and seedream_get_tasks_batch.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides explicit use cases (checking completion, retrieving URLs, getting full details) but does not mention when not to use or compare to seedream_get_tasks_batch.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Query multiple Seedream image tasks at once.
Efficiently check the status of multiple tasks in a single request.
More efficient than calling seedream_get_task multiple times.
Use this when:
- You have multiple pending generations to check
- You want to get status of several images at once
- You're tracking a batch of generations
Returns:
Status and image information for all queried tasks.
| Name | Required | Description | Default |
|---|---|---|---|
| task_ids | Yes | List of task IDs to query. Allows querying multiple tasks at once. |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries full burden. It describes the operation as querying (read-only) and states it returns 'Status and image information'. It does not mention auth requirements or rate limits, but for a query tool, this is sufficient. Slightly more detail on the response format would improve, but overall transparent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise with a clear initial sentence, a bullet-style list for usage scenarios, and a returns section. No redundant information; every sentence adds value.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the simplicity (1 parameter, no annotations, output schema exists), the description sufficiently covers purpose, usage, and returns. It doesn't explain output schema details, but that's acceptable as the schema itself is present. A hint about pagination or result format could improve, but not necessary.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% with a clear parameter description. The tool description adds context about efficiency and use cases but does not add new parameter semantics beyond what the schema provides. Baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states 'Query multiple Seedream image tasks at once' and distinguishes from the sibling tool 'seedream_get_task' by highlighting batch efficiency. The verb 'query' and resource 'tasks' are specific.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides explicit when-to-use scenarios: 'when you have multiple pending generations', 'when you want to get status of several images at once', 'when tracking a batch'. Although not explicitly stating when not to use, it implies the alternative for single tasks (seedream_get_task).
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List all available Seedream models with their capabilities and pricing.
Use this when:
- User asks what models are available
- You need to help choose the right model for a task
- You want to compare model capabilities
Returns:
Formatted table of all Seedream models with descriptions.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations provided, so description carries full burden. It states it returns a formatted table with descriptions, implying no side effects. Does not mention authentication requirements or data freshness, but for a read-only listing tool it is adequately transparent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Concise: one line for purpose, bulleted use cases, and output format. No fluff. Well-structured and easy to scan.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given zero parameters and an output schema exists, description is complete. Explains what models are listed and for what purpose. Does not need to explain return values since output schema covers that.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
No parameters in input schema. Description compensates by detailing what the output includes (capabilities, pricing, descriptions). Baseline 4 for zero-param tools where description adds meaningful output context.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description clearly states it lists all available Seedream models with capabilities and pricing. Specific verb 'list' and resource 'models'. Distinguishes from sibling tools which focus on editing, generating, or retrieving tasks.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly lists three use cases: user asks about models, need to choose a model, want to compare capabilities. Does not mention when not to use, but that is not critical for a simple listing tool. No alternative tools suggested, but sibling tools are distinct in purpose.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List all available image sizes and resolution options for Seedream.
Use this when:
- User asks about available image sizes
- You need to help choose the right resolution
- You want to understand size options
Returns:
Formatted list of all size options with descriptions.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It states it returns a 'formatted list of all size options with descriptions', which is sufficient for a simple read-only listing tool with no parameters. It does not disclose potential rate limits or side effects, but these are minimal for such a tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise, with a clear 'Use this when' section and a 'Returns' section. Every sentence adds value, and the structure is front-loaded with the core purpose.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool has zero parameters and a simple list output, the description fully covers its purpose, usage context, and return value. The output schema exists, and the description notes the return format adequately.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool has no parameters, and schema description coverage is 100%. Per guidelines, baseline is 3 when coverage is high. The description does not add any parameter semantics since there are none to explain.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb 'List' and resource 'image sizes and resolution options for Seedream'. It clearly distinguishes from sibling tools like seedream_generate_image and seedream_list_models.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description includes a 'Use this when' section listing three clear scenarios. While it doesn't explicitly state when not to use it or mention alternatives, the guidelines are adequate for an agent to understand when to invoke this tool.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.16seedream_edit_image2 fields changedInput schema / properties / model / defaultPrevious value: -"doubao-seedream-5-0-260128"New value: +"doubao-seedream-5-0-lite-260128"Input schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-pro-260628",
- "doubao-seedream-5-0-260128",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828"
-]New value: +[
+ "doubao-seedream-5-0-pro-260628",
+ "doubao-seedream-5-0-lite-260128",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828"
+]seedream_generate_image4 fields changedInput schema / properties / model / defaultPrevious value: -"doubao-seedream-5-0-260128"New value: +"doubao-seedream-5-0-lite-260128"Input schema / properties / model / descriptionPrevious value: -"Model to use for generation. 'doubao-seedream-5-0-pro-260628' (v5.0 Pro, flagship single image, highest quality; no sequential generation, streaming, or web search). 'doubao-seedream-5-0-260128' (v5.0 Lite, latest flagship, sequential generation, streaming, web search). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value)."New value: +"Model to use for generation. 'doubao-seedream-5-0-pro-260628' (v5.0 Pro, flagship single image, highest quality; no sequential generation, streaming, or web search). 'doubao-seedream-5-0-lite-260128' (v5.0 Lite, sequential generation, streaming, web search). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value)."Input schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-pro-260628",
- "doubao-seedream-5-0-260128",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828"
-]New value: +[
+ "doubao-seedream-5-0-pro-260628",
+ "doubao-seedream-5-0-lite-260128",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828"
+]Input schema / properties / tools / descriptionPrevious value: -"Optional list of tool types for the model to use during generation. Currently only 'web_search' is supported. Only supported by doubao-seedream-5-0-260128 (v5.0)."New value: +"Optional list of tool types for the model to use during generation. Currently only 'web_search' is supported. Only supported by doubao-seedream-5-0-lite-260128 (v5.0 Lite)."v0.1.15seedream_decompose_imageseedream_edit_image5 fields changedInput schema / properties / backgroundAdded value: +{
+ "anyOf": [
+ {
+ "enum": [
+ "transparent",
+ "opaque"
+ ],
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Seedream 5.0 Pro background mode. transparent requires one PNG input and PNG output.",
+ "title": "Background"
+}Input schema / properties / optimize_prompt_options / descriptionPrevious value: -"Optional prompt optimization configuration."New value: +"Prompt optimization. Pro supports standard/fast; Lite supports standard."Input schema / properties / size / anyOfPrevious value: -[
- {
- "enum": [
- "1K",
- "2K",
- "3K",
- "4K"
- ],
- "type": "string"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+]Input schema / properties / size / descriptionPrevious value: -"Output image resolution. '1K' (default), '2K', '3K', or '4K'."New value: +"Model-specific output size or explicit dimensions. Pro supports 1K/1.5K/2K; Lite supports 2K/3K/4K."Input schema / properties / streamRemoved value: -{
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Stream pictures progressively when supported.",
- "title": "Stream"
-}seedream_generate_image5 fields changedInput schema / properties / sequential_image_generation / descriptionPrevious value: -"Generate related images based on input. 'auto' enables it, 'disabled' (default) turns it off. Only supports v4.5 and v4.0 models."New value: +"Generate related images based on input. 'auto' enables it, 'disabled' (default) turns it off. Supported by Seedream 5.0 Lite, 4.5, and 4.0; not Pro."Input schema / properties / sequential_image_generation_options / descriptionPrevious value: -"Tunable options for grouped image generation. Only honored when `sequential_image_generation=auto`. Supports `max_images` (int, range [1, 15]). Only supported on doubao-seedream-4.5 and doubao-seedream-4.0."New value: +"Tunable options for grouped image generation. Only honored when `sequential_image_generation=auto`. Supports `max_images` (int, range [1, 15]). Supported by Seedream 5.0 Lite, 4.5, and 4.0; not Pro."Input schema / properties / size / anyOfPrevious value: -[
- {
- "enum": [
- "1K",
- "2K",
- "3K",
- "4K"
- ],
- "type": "string"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+]Input schema / properties / size / descriptionPrevious value: -"Output image resolution. '1K' (default), '2K', '3K', or '4K'. You can also specify custom dimensions like '1024x1024', '1280x720', etc."New value: +"Model-specific output size: Pro supports 1K/1.5K/2K (and auto for decomposition); Lite supports 2K/3K/4K. Explicit dimensions such as 2048x1024 are also accepted."Input schema / properties / streamRemoved value: -{
- "anyOf": [
- {
- "type": "boolean"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Stream all pictures progressively. Default is false. Only supports v4.5 and v4.0 models.",
- "title": "Stream"
-}v0.1.14seedream_edit_image5 fields changedInput schema / properties / optimize_prompt_optionsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": true,
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Optional prompt optimization configuration.",
+ "title": "Optimize Prompt Options"
+}Input schema / properties / sequential_image_generationAdded value: +{
+ "anyOf": [
+ {
+ "enum": [
+ "auto",
+ "disabled"
+ ],
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Generate related images based on input. 'auto' enables it.",
+ "title": "Sequential Image Generation"
+}Input schema / properties / sequential_image_generation_optionsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": true,
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Tunable options for grouped image generation.",
+ "title": "Sequential Image Generation Options"
+}Input schema / properties / streamAdded value: +{
+ "anyOf": [
+ {
+ "type": "boolean"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Stream pictures progressively when supported.",
+ "title": "Stream"
+}Input schema / properties / toolsAdded value: +{
+ "anyOf": [
+ {
+ "items": {
+ "const": "web_search",
+ "type": "string"
+ },
+ "type": "array"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Optional list of tool types for the model to use during editing.",
+ "title": "Tools"
+}v0.1.11seedream_edit_image2 fields changedInput schema / properties / size / anyOfPrevious value: -[
- {
- "enum": [
- "1K",
- "2K",
- "3K",
- "4K",
- "adaptive"
- ],
- "type": "string"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "enum": [
+ "1K",
+ "2K",
+ "3K",
+ "4K"
+ ],
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+]Input schema / properties / size / descriptionPrevious value: -"Output image resolution. '1K' (default), '2K', '3K', '4K', or 'adaptive'."New value: +"Output image resolution. '1K' (default), '2K', '3K', or '4K'."seedream_generate_image2 fields changedInput schema / properties / size / anyOfPrevious value: -[
- {
- "enum": [
- "1K",
- "2K",
- "3K",
- "4K",
- "adaptive"
- ],
- "type": "string"
- },
- {
- "type": "null"
- }
-]New value: +[
+ {
+ "enum": [
+ "1K",
+ "2K",
+ "3K",
+ "4K"
+ ],
+ "type": "string"
+ },
+ {
+ "type": "null"
+ }
+]Input schema / properties / size / descriptionPrevious value: -"Output image resolution. '1K' (default), '2K', '3K', '4K', or 'adaptive'. You can also specify custom dimensions like '1024x1024', '1280x720', etc."New value: +"Output image resolution. '1K' (default), '2K', '3K', or '4K'. You can also specify custom dimensions like '1024x1024', '1280x720', etc."v0.1.10seedream_edit_image5 fields changedInput schema / properties / guidance_scaleRemoved value: -{
- "anyOf": [
- {
- "type": "number"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Prompt weight — higher values make edits follow the prompt more closely. Range: [1, 10]. Default is 5.5 for doubao-seededit-3-0-i2i. Only works with v3 models.",
- "title": "Guidance Scale"
-}Input schema / properties / model / defaultPrevious value: -"doubao-seededit-3-0-i2i-250628"New value: +"doubao-seedream-5-0-260128"Input schema / properties / model / descriptionPrevious value: -"Model to use for editing. 'doubao-seededit-3-0-i2i-250628' (dedicated editing model, best for image modification). Other models can also be used for editing when images are provided."New value: +"Model to use for editing. Seedream 5.0 Pro, 5.0 Lite, 4.5, and 4.0 all support image editing when images are provided."Input schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-pro-260628",
- "doubao-seedream-5-0-260128",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828",
- "doubao-seedream-3-0-t2i-250415",
- "doubao-seededit-3-0-i2i-250628"
-]New value: +[
+ "doubao-seedream-5-0-pro-260628",
+ "doubao-seedream-5-0-260128",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828"
+]Input schema / properties / seedRemoved value: -{
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Random seed for reproducible edits. Range: [-1, 2147483647]. Default is -1 (random). Only works with v3 models.",
- "title": "Seed"
-}seedream_generate_image4 fields changedInput schema / properties / guidance_scaleRemoved value: -{
- "anyOf": [
- {
- "type": "number"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Prompt weight — higher values make the result more closely follow the prompt. Range: [1, 10]. Default is 2.5 for doubao-seedream-3-0-t2i. Only works with v3 models.",
- "title": "Guidance Scale"
-}Input schema / properties / model / descriptionPrevious value: -"Model to use for generation. 'doubao-seedream-5-0-pro-260628' (v5.0 Pro, flagship single image, highest quality; no sequential generation, streaming, or web search). 'doubao-seedream-5-0-260128' (v5.0 Lite, latest flagship, sequential generation, streaming, web search). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value). 'doubao-seedream-3-0-t2i-250415' (v3 text-to-image, supports seed and guidance_scale). 'doubao-seededit-3-0-i2i-250628' is for image editing only — use seedream_edit_image instead."New value: +"Model to use for generation. 'doubao-seedream-5-0-pro-260628' (v5.0 Pro, flagship single image, highest quality; no sequential generation, streaming, or web search). 'doubao-seedream-5-0-260128' (v5.0 Lite, latest flagship, sequential generation, streaming, web search). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value)."Input schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-pro-260628",
- "doubao-seedream-5-0-260128",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828",
- "doubao-seedream-3-0-t2i-250415",
- "doubao-seededit-3-0-i2i-250628"
-]New value: +[
+ "doubao-seedream-5-0-pro-260628",
+ "doubao-seedream-5-0-260128",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828"
+]Input schema / properties / seedRemoved value: -{
- "anyOf": [
- {
- "type": "integer"
- },
- {
- "type": "null"
- }
- ],
- "default": null,
- "description": "Random seed for reproducible results. Range: [-1, 2147483647]. Default is -1 (random). Only works with v3 models (doubao-seedream-3-0-t2i and doubao-seededit-3-0-i2i).",
- "title": "Seed"
-}v0.1.9seedream_edit_image1 field changedInput schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-260128",
- "doubao-seedream-5.0-lite",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828",
- "doubao-seedream-3-0-t2i-250415",
- "doubao-seededit-3-0-i2i-250628"
-]New value: +[
+ "doubao-seedream-5-0-pro-260628",
+ "doubao-seedream-5-0-260128",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828",
+ "doubao-seedream-3-0-t2i-250415",
+ "doubao-seededit-3-0-i2i-250628"
+]seedream_generate_image2 fields changedInput schema / properties / model / descriptionPrevious value: -"Model to use for generation. 'doubao-seedream-5-0-260128' (v5.0, latest flagship, highest quality). 'doubao-seedream-5.0-lite' (v5.0 economy variant, faster and lower cost). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value). 'doubao-seedream-3-0-t2i-250415' (v3 text-to-image, supports seed and guidance_scale). 'doubao-seededit-3-0-i2i-250628' is for image editing only — use seedream_edit_image instead."New value: +"Model to use for generation. 'doubao-seedream-5-0-pro-260628' (v5.0 Pro, flagship single image, highest quality; no sequential generation, streaming, or web search). 'doubao-seedream-5-0-260128' (v5.0 Lite, latest flagship, sequential generation, streaming, web search). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value). 'doubao-seedream-3-0-t2i-250415' (v3 text-to-image, supports seed and guidance_scale). 'doubao-seededit-3-0-i2i-250628' is for image editing only — use seedream_edit_image instead."Input schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-260128",
- "doubao-seedream-5.0-lite",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828",
- "doubao-seedream-3-0-t2i-250415",
- "doubao-seededit-3-0-i2i-250628"
-]New value: +[
+ "doubao-seedream-5-0-pro-260628",
+ "doubao-seedream-5-0-260128",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828",
+ "doubao-seedream-3-0-t2i-250415",
+ "doubao-seededit-3-0-i2i-250628"
+]v0.1.8seedream_edit_image1 field changedInput schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-260128",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828",
- "doubao-seedream-3-0-t2i-250415",
- "doubao-seededit-3-0-i2i-250628"
-]New value: +[
+ "doubao-seedream-5-0-260128",
+ "doubao-seedream-5.0-lite",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828",
+ "doubao-seedream-3-0-t2i-250415",
+ "doubao-seededit-3-0-i2i-250628"
+]seedream_generate_image2 fields changedInput schema / properties / model / descriptionPrevious value: -"Model to use for generation. 'doubao-seedream-5-0-260128' (v5.0, latest flagship, highest quality). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value). 'doubao-seedream-3-0-t2i-250415' (v3 text-to-image, supports seed and guidance_scale). 'doubao-seededit-3-0-i2i-250628' is for image editing only — use seedream_edit_image instead."New value: +"Model to use for generation. 'doubao-seedream-5-0-260128' (v5.0, latest flagship, highest quality). 'doubao-seedream-5.0-lite' (v5.0 economy variant, faster and lower cost). 'doubao-seedream-4-5-251128' (v4.5, previous flagship, great quality). 'doubao-seedream-4-0-250828' (v4.0, stable, best value). 'doubao-seedream-3-0-t2i-250415' (v3 text-to-image, supports seed and guidance_scale). 'doubao-seededit-3-0-i2i-250628' is for image editing only — use seedream_edit_image instead."Input schema / properties / model / enumPrevious value: -[
- "doubao-seedream-5-0-260128",
- "doubao-seedream-4-5-251128",
- "doubao-seedream-4-0-250828",
- "doubao-seedream-3-0-t2i-250415",
- "doubao-seededit-3-0-i2i-250628"
-]New value: +[
+ "doubao-seedream-5-0-260128",
+ "doubao-seedream-5.0-lite",
+ "doubao-seedream-4-5-251128",
+ "doubao-seedream-4-0-250828",
+ "doubao-seedream-3-0-t2i-250415",
+ "doubao-seededit-3-0-i2i-250628"
+]v0.1.5seedream_edit_imageseedream_generate_imageseedream_get_taskseedream_get_tasks_batchseedream_list_modelsseedream_list_sizesv0.1.3seedream_edit_imageseedream_generate_imageseedream_get_taskseedream_get_tasks_batchseedream_list_modelsseedream_list_sizesv0.1.1seedream_generate_image2 fields changedInput schema / properties / optimize_prompt_optionsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": true,
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Optional prompt optimization configuration. Supports `mode` with values 'standard' (higher quality, slower) or 'fast' (quicker, lower quality). Only supported on doubao-seedream-4.5 (standard mode only) and doubao-seedream-4.0.",
+ "title": "Optimize Prompt Options"
+}Input schema / properties / sequential_image_generation_optionsAdded value: +{
+ "anyOf": [
+ {
+ "additionalProperties": true,
+ "type": "object"
+ },
+ {
+ "type": "null"
+ }
+ ],
+ "default": null,
+ "description": "Tunable options for grouped image generation. Only honored when `sequential_image_generation=auto`. Supports `max_images` (int, range [1, 15]). Only supported on doubao-seedream-4.5 and doubao-seedream-4.0.",
+ "title": "Sequential Image Generation Options"
+}v0.1.0seedream_edit_imageseedream_generate_imageseedream_get_taskseedream_get_tasks_batchseedream_list_modelsseedream_list_sizesScored across 7 tools
Each tool targets a clearly distinct operation: generate (text-to-image), edit (modify existing), decompose (split into layers), and the query/introspection tools. The single-vs-batch task querying pair (seedream_get_task vs seedream_get_tasks_batch) has an explicitly documented boundary, and list_models vs list_sizes cover different concerns.
Every tool follows a uniform seedream_ prefix plus a consistent verb_noun snake_case pattern (generate_image, edit_image, get_task, list_models). No mixing of conventions or casing styles.
Seven tools is well-scoped for an image generation service, with no redundant entries. Each tool maps to a distinct capability (three image operations, two query tools, two discovery tools).
Core lifecycle is covered: image creation, editing, decomposition, task status polling (single and batch), plus model and size discovery. Minor gaps exist around task management actions (e.g. cancel/delete task) and post-processing like upscaling, but the main workflows are fully supported.
AI video, images, music & SFX: Seedance 2.5, Veo 3.1, Kling 3.0, Nano Banana Pro, 20+ models.
Generate images, video, music, voice and 3D through one API. 30 tools, 200+ models.
Best Image and video generation: 20+ models (Kling, Seedance, Veo, NB, FLUX.2), OAuth, pay-per-use.
MCP server for ByteDance Seedance AI video generation
