Skip to main content
Glama
justForever17

jimeng_visual_generation

by justForever17
OverviewSchemaRelated ServersScoreDiscussions
Python
Remote

Jimeng 视觉生成 MCP 服务器

基于 Volcengine（火山引擎）Jimeng 视觉生成 API 的 MCP 服务器。通过模型上下文协议（MCP）提供图像和视频生成功能。

功能

  • 文本生成图像（T2I）：根据文本提示生成图像

  • 图像到图像（I2I）：基于提示和参考图像对图像进行变换

  • 多图融合：合成多张参考图像

  • 文本生成视频（T2V）：根据文本提示生成视频

  • 图像到视频（I2V）：从首帧或首帧+末帧图像生成视频

  • 自动本地文件支持：可传入图像公网URL，同时支持本地文件路径，程序会自动将其转换为 Base64

Related MCP server: jimeng4-mcp

安装

选项 1：通过 pip 安装

pip install jimeng_visual_generation

选项 2：直接使用 uvx 运行（推荐）

无需安装。uvx 会自动下载并运行该包：

uvx jimeng_visual_generation

在 VS Code / Cursor / Claude Desktop 中的配置

将下列内容添加到你的 MCP 配置文件：

  • VS Code：~/.vscode/mcp.json 或 工作区设置

  • Cursor：Settings -> MCP Servers

  • Claude Desktop：%APPDATA%\Claude\claude_desktop_config.json

示例配置（使用环境变量）

{
  "mcpServers": {
    "jimeng_visual_generation": {
      "command": "uvx",
      "args": ["jimeng_visual_generation"],
      "env": {
        "VOLC_API_KEY": "your_volcengine_api_key_here",
        "VOLC_IMAGE_MODEL": "doubao-seedream-5-0-pro-260628",
        "VOLC_VIDEO_MODEL": "doubao-seedance-2-5-260628"
      }
    }
  }
}

环境变量

变量

是否必需

描述

VOLC_API_KEY

✅ 必需

你的 Volcengine API Key

VOLC_IMAGE_MODEL

可选

图像模型 ID（默认：doubao-seedream-5-0-pro-260628）

VOLC_VIDEO_MODEL

可选

视频模型 ID（默认：doubao-seedance-2-5-260628）

可用工具

generate_image

使用文本提示和可选参考图像生成图像。支持最新的 Doubao-Seedream-5.0-pro 精准图层编辑。

参数：

  • prompt（必需）：描述目标图像的文本

  • image_urls（可选）：参考图像列表（支持 URL、Base64 或本地文件路径）

  • model（可选）：使用的模型 ID 或 Endpoint ID（例如 doubao-seedream-5-0-pro-260628）

  • size（可选）：图像尺寸比例（支持 "1:1", "16:9", "2K", "4K" 等，禁止使用 ratio）

generate_video

创建视频生成任务。支持多种生成模式（包括最新的 Seedance 2.5 30秒超长叙事与全模态参考）：

  • 文本生成视频 (T2V)：不提供图像、视频、音频输入，只提供 prompt。

  • 首帧/尾帧生视频 (I2V)：提供 1-2 张图像。

  • 多模态参考生视频 (Seedance 2.5 / 2.0 / 2.0-mini)：可混合提供图像（最多9张）、参考视频（最多3个）和参考音频（最多3个），支持本地文件自动转换。

  • 30秒超长叙事 (Seedance 2.5)：最高支持 30 秒长视频生成。

参数：

  • prompt（可选）：视频描述的文本提示词

  • image_urls（可选）：输入参考图像列表（支持 URL、Base64 或本地文件路径，最多 9 张）

  • video_urls（可选）：参考视频列表（支持 URL 或本地文件路径，最多 3 个，Seedance 2.5 支持 ≤ 30s，2.0 支持 ≤ 15s）

  • audio_urls（可选）：参考音频列表（支持 URL 或本地文件路径，最多 3 个）

  • image_roles（可选）：为 image_urls 显式指定的角色列表（例如 ["reference_image", "first_frame"]）

  • model（可选）：使用的模型 ID 或 Endpoint ID（如 doubao-seedance-2-5-260628、doubao-seedance-2-0-mini-260615 等）

  • ratio（可选）：宽高比（例如："16:9"、"9:16"，禁止在图片生成里使用该参数）

  • resolution（可选）：分辨率（"720p" 或 "1080p"，mini 与 fast 模型支持至 720p，2.5 与标准 2.0 支持 1080p）

  • duration（可选）：视频时长（秒，Seedance 2.5 支持 4-30s，Seedance 2.0/2.0-mini 支持 4-15s，默认 5）

  • return_last_frame（可选）：是否返回生成的视频最后一帧图像 URL，适用于连续生成

get_video_task_result

查询视频生成任务的状态和结果。

参数：

  • task_id（必需）：由 generate_video 返回的任务 ID

📄 开源协议 (License)

本项目基于 MIT License 开源。

💖 赞助 (Sponsorship)

维护开源项目不易，如果您觉得 jimeng_visual_generation 对您有帮助，欢迎请作者喝杯咖啡！

平台

链接

支付方式

爱发电 (Afdian)

Afdian

微信, 支付宝

Available Tools

3 tools
generate_imageA

Generate images using Volcengine visual generation API.

CRITICAL INSTRUCTIONS FOR AI AGENTS:

  • For image size, ONLY use the 'size' parameter (e.g., size="16:9"). DO NOT pass a 'ratio' or 'width'/'height' parameter.

  • If the user provides a custom Endpoint ID (starts with 'ep-'), you MUST pass it into the 'model' parameter.

Capabilities:

  1. Text-to-Image: Provide 'prompt' and 'size'.

  2. Image-to-Image: Provide 'prompt' AND 'image_urls' (list containing 1 image string).

ParametersJSON Schema
NameRequiredDescriptionDefault
paramsYes

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.7/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations exist, so the description carries full burden. It discloses critical parameter behaviors (e.g., 'model' should be omitted unless custom endpoint) and mode-specific requirements. However, it does not mention rate limits, authentication, or idempotency, leaving some gaps.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is well-structured with sections for critical instructions and capabilities. It is efficient but slightly verbose in the instructions section; however, every sentence serves a purpose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's complexity (two modes, multiple constraints), the description covers essential usage details. The presence of an output schema reduces the need to describe return values, making the description complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description adds significant context beyond the input schema, such as the prohibition of 'ratio' and the rule for 'model'. It also explains the relationship between parameters for different modes, enhancing understanding of parameter usage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool generates images using a specific API. It lists two capabilities (text-to-image and image-to-image), and sibling tools ('generate_video', 'get_video_task_result') are distinct, making the purpose unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit instructions: use 'size' parameter only, avoid 'ratio'/'width'/'height', and pass custom endpoint IDs to 'model'. It also explains when to use each mode based on provided parameters, offering clear guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

generate_videoA

Create a video generation task using Volcengine API.

CRITICAL INSTRUCTIONS FOR AI AGENTS:

  • This tool ONLY submits the task. It returns a Task ID. You MUST subsequently call get_video_task_result in a loop (wait 5-10s between calls) to retrieve the actual video URL.

  • Use 'ratio' for video proportions (e.g., ratio="16:9"). DO NOT use 'size' or 'width'/'height'.

  • If user provides an Endpoint ID ('ep-...'), pass it to the 'model' parameter.

Capabilities:

  1. Text-to-Video: Provide 'prompt' and 'ratio'.

  2. Image-to-Video: Provide 'image_urls' array + optional 'prompt'.

  3. Multi-Modal (Seedance 2.0): Mix 'image_urls' (up to 9), 'video_urls' (up to 3), and 'audio_urls' (up to 3).

ParametersJSON Schema
NameRequiredDescriptionDefault
paramsYes

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A5/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Without annotations, the description fully discloses behavior: asynchronous (returns Task ID, needs polling), default values, accepted values for ratio and resolution, automatic role assignment for images, and model-specific limitations (e.g., generate_audio only for Pro models).

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Well-structured: starts with main purpose, then critical instructions in bold, followed by capabilities. Each sentence adds value without redundancy. Uses formatting for emphasis and lists for clarity.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Fully covers all aspects needed for correct usage: explains the asynchronous workflow, all parameter constraints, mode-specific behaviors, and edge cases (e.g., seed for reproducibility, watermark option). With an output schema present, the description appropriately focuses on input semantics and process.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Adds critical meaning beyond the input schema: clarifies that 'model' should be omitted unless user specifies an Endpoint ID, that 'ratio' must not use 'size' parameters, and explains the roles of image_urls, video_urls, and audio_urls for different modes. Enumerates accepted values for ratio explicitly.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Explicitly states "Create a video generation task" and distinguishes from sibling tools by noting that this tool only submits the task and returns a Task ID, while `get_video_task_result` must be polled later.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Provides clear when-to-use instructions: only for submitting tasks, not for retrieving results. Gives detailed do's and don'ts (e.g., use 'ratio' not 'size', leave 'model' empty unless custom endpoint). Lists three capability modes with their parameter requirements.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

get_video_task_resultA

Query the status of a video generation task using the Task ID returned by generate_video.

CRITICAL INSTRUCTIONS FOR AI AGENTS:

  • Video generation takes time (often 1-3 minutes).

  • If the returned status is "ordered" or "running", DO NOT tell the user it failed. Ask the user to wait, and call this tool again after 10-15 seconds.

  • Once the status is "succeeded", the response will contain the download URL.

ParametersJSON Schema
NameRequiredDescriptionDefault
paramsYes

Output Schema

ParametersJSON Schema
NameRequiredDescription
resultYes

TDQS

A4.3/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description fully bears the load. It discloses polling behavior, expected duration (1-3 minutes), and appropriate agent actions. Does not mention potential failures beyond statuses.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Concise, well-structured with clear sections including critical instructions. Every sentence adds value, no redundancy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the output schema exists (not provided but indicated), the description sufficiently covers the polling loop and status handling. Complete for a query tool with one parameter.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Only one parameter (task_id) which is well-described in the schema itself ('Task ID from generate_video'). The description adds no additional meaning beyond the schema's property description, so baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Clearly states it queries the status of a video generation task using the Task ID from generate_video. Distinguishes from sibling tools generate_image and generate_video by focusing on result retrieval.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Provides explicit instructions on when to call this tool (after generate_video) and how to handle different statuses (ordered/running vs succeeded). Lacks explicit 'when not to use' but is clear enough.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. 3 tool updatesv0.2.2
    • First observedgenerate_image
    • First observedgenerate_video
    • First observedget_video_task_result
Install Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

A4.4/5.0
Disambiguation5/5

Each tool targets a distinct action: image generation, video submission, and video result polling. No overlap in functionality.

Naming Consistency5/5

All tools follow a consistent verb_noun pattern in snake_case: generate_image, generate_video, get_video_task_result.

Tool Count4/5

Three tools is minimal but covers the core workflows of image generation and video generation with polling. Slightly thin but reasonable for a focused server.

Completeness3/5

Missing operations like listing or deleting generated content, and no image editing or video task cancellation. Gaps exist but essential flows are present.

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Connectors

Related MCP Servers

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/justForever17/jimeng_visual_generation'

If you have feedback or need assistance with the MCP directory API, please join our Discord server