Mimo Vision MCP
Provides image recognition capabilities using Xiaomi's Mimo vision API, including image description, OCR, and structured data extraction.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Mimo Vision MCPExtract text from screenshot.png"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Mimo Vision MCP
基于 mimo-v2.5 的图像识别 MCP 服务器,为 AI 助手提供"看图"能力。
功能特性
单一工具 recognize_image,通过 mode 参数切换三种识别能力:
模式 | 用途 | 输出格式 | 典型场景 |
| 通用图像描述 | 纯文本 | 场景理解、内容审核、图片摘要 |
| 文字识别 | 纯文本 | 截图提取、文档数字化、车牌/招牌识别 |
| 结构化信息提取 | JSON | 发票解析、名片录入、表格数据化 |
extract 模式输出格式
通过 format 参数控制输出结构:
值 | 行为 |
| 由模型自动判断最合适的结构 |
| 强制输出 JSON 对象 |
| 强制输出 Markdown 表格 |
Related MCP server: image-analysis-mcp
快速开始
1. 安装
# 克隆仓库
git clone https://github.com/k0tori/mimo-vision-mcp.git
cd mimo-vision-mcp
# 安装(开发模式)
pip install -e .2. 配置 API
cp .env.example .env编辑 .env,填入你的 API 凭据:
MIMO_API_BASE_URL=https://your-api-endpoint/v1
MIMO_API_KEY=your-api-key3. 注册 MCP 服务器
在你的 MCP 客户端配置中添加:
{
"mcpServers": {
"mimo-vision": {
"command": "python",
"args": ["-m", "mimo_vision_mcp"],
"env": {
"MIMO_API_BASE_URL": "https://your-api-endpoint/v1",
"MIMO_API_KEY": "your-api-key"
}
}
}
}提示:
env字段可选。如果省略,服务器会从项目目录下的.env文件读取配置。
使用示例
describe — 描述图片内容
{
"image": "/path/to/photo.jpg",
"mode": "describe"
}也支持 URL 输入:
{ "image": "https://example.com/photo.jpg", "mode": "describe" }
ocr — 提取文字
{
"image": "/path/to/screenshot.png",
"mode": "ocr"
}extract — 结构化提取
{
"image": "/path/to/invoice.jpg",
"mode": "extract",
"prompt": "提取发票号码、开票日期和总金额"
}强制 JSON 输出:
{
"image": "/path/to/business-card.png",
"mode": "extract",
"format": "json"
}工具接口
recognize_image
参数 | 类型 | 必填 | 默认值 | 说明 |
| string | ✅ | — | 本地文件路径或图片 URL |
| string | ✅ | — |
|
| string | — |
| 补充提示词,引导模型关注特定内容 |
| string | — |
| 仅 extract 模式: |
支持的图片格式
格式 | 扩展名 | 最大大小 |
PNG |
| 20 MB |
JPEG |
| 20 MB |
WebP |
| 20 MB |
GIF |
| 20 MB |
格式通过文件头(magic bytes)自动检测,不依赖扩展名。
错误处理
所有错误以纯文本返回,不会抛出异常。常见错误类型:
错误 | 示例 |
文件不存在 |
|
格式不支持 |
|
文件过大 |
|
API 认证失败 |
|
模型返回异常 |
|
开发
# 安装开发依赖
pip install -e ".[dev]"
# 运行测试
pytest
# 运行测试(详细输出)
pytest -v项目结构
src/mimo_vision_mcp/
├── __init__.py # 包初始化
├── __main__.py # python -m 入口
├── server.py # MCP 服务,注册工具
├── api/
│ └── client.py # OpenAI 兼容 API 客户端
├── tools/
│ └── recognize.py # recognize_image 实现
└── utils/
├── image.py # 图片处理(magic bytes、base64)
└── prompts.py # System prompt 定义License
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for Qwen Image 3 AI image generation
OCR, transcription, file extraction, and image generation for AI agents via MCP.
Hosted MCP server for AI photo editing: nothing to install, no desktop photo editor needed.
MCP server for Midjourney AI image generation and editing
Related MCP Servers
- AlicenseAqualityDmaintenanceMCP server for image recognition, supporting multiple vision backends (Anthropic, Zhipu, Ollama) to describe, answer questions, and analyze images.318 npm1MIT
- FlicenseAqualityCmaintenanceMCP server for image analysis, providing OCR text extraction, image metadata, and EXIF/GPS data as structured JSON for downstream processing.3-

mimo-vision-mcpofficial
AlicenseAqualityCmaintenanceAn MCP server that uses Xiaomi MiMo v2.5 multimodal model to provide image recognition capabilities (description, multi-image analysis, OCR, and image info validation) for text-only main models like deepseek-v4-flash, accepting local paths, URLs, file://, and base64 data inputs.41MIT- AlicenseBqualityCmaintenanceMCP server that provides image analysis, OCR text extraction, and image description using the GLM-4V Flash model from Zhipu AI.38MIT