F
licenseB
qualityD
maintenanceEnables intelligent image analysis using GLM-4.5V model, specializing in extracting and analyzing code from screenshots with support for file paths and clipboard input.
1
11
-
Vision OCR MCP
by wbwbwb123
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., "@Vision OCR MCPOCR my clipboard image"
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.
Vision OCR MCP
本地 RapidOCR（rapidocr + ONNX Runtime，默认免费）+ 通义千问 Qwen-VL（summarize / both）。
能力
mode | 说明 |
| 本地 OCR，中英截图/报错 |
| 通义千问视觉总结（需 API Key） |
| OCR 原文 + 千问总结 |
输入：path 本地路径，或 from_clipboard=true 剪贴板图片。不落盘，结果直接返回给 Agent。
Related MCP server: image-recognition-mcp
安装
cd d:\code\cuosor\skills+mcp\vision-ocr-mcp
python -m pip install -e .
copy .env.example .env
# 编辑 .env，填入 DASHSCOPE_API_KEY（仅 summarize/both 需要）接入 Cursor
编辑用户 MCP 配置（常见路径）：%USERPROFILE%\.cursor\mcp.json
{
"mcpServers": {
"vision-ocr": {
"command": "D:\\Anaconda3-new\\python.exe",
"args": ["-m", "vision_ocr_mcp"],
"cwd": "d:\\code\\cuosor\\skills+mcp\\vision-ocr-mcp",
"env": {
"DASHSCOPE_API_KEY": "sk-你的密钥",
"QWEN_VL_MODEL": "qwen-vl-plus",
"PYTHONPATH": "d:\\code\\cuosor\\skills+mcp\\vision-ocr-mcp\\src"
}
}
}
}若已 pip install -e .，可去掉 PYTHONPATH，command 用安装了该包的同一个 Python。
改完后重启 Cursor / 重载 MCP。对话里可说：
「用 vision-ocr 读这张图
D:\xxx.png」「从剪贴板 OCR」
「用 summarize 总结这张架构图」
费用
OCR：本地，免费
千问：按 Token；新用户常有免费额度，见阿里云百炼控制台
云端高精度 OCR 未接入，可后续再加。
This server cannot be deployed
Maintenance
ActivityInactive
ResponsivenessNo issues
Related MCP Connectors
OCR and document understanding: extract text from images, then summarize or translate it.
OCR, transcription, file extraction, and image generation for AI agents via MCP.
Local ML: speech to text (SRT/VTT), text embeddings, reranking, similarity, entities, zero-shot.
PDF, image, video, OCR, screenshot, SQL, QR and text tools for agents. No API key, no signup.
Related MCP Servers
- AlicenseAqualityDmaintenanceGives vision-less LLMs the ability to recognize clipboard screenshots and images by proxying to an OpenAI-compatible vision model.116 npm2MIT
- AlicenseCqualityDmaintenanceA local OCR MCP server that extracts text from images using PP-OCRv6 for fast text extraction and VL-1.6 for document structure analysis, with automatic model routing and GPU detection.32MIT
- AlicenseNot gradedqualityCmaintenanceMCP server that adds Qwen3-VL vision capabilities to Claude Code or any MCP client, enabling OCR, UI/screenshot recognition, and chart understanding. It provides tools for analyzing single or batch images via DashScope's OpenAI-compatible API, with caching and fast mode.MIT