MiniCPM Vision MCP
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., "@MiniCPM Vision MCP描述这张图片 C:\照片\example.jpg"
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.
MiniCPM Vision MCP 🎬👁️🎵
为 DeepSeek V4.0 等单模态大模型 装上眼睛和耳朵 —— 本地运行的视觉+音频 MCP 服务器,基于 Ollama + MiniCPM-V 4.6 + faster-whisper。
图片描述 · 视频帧分析 · 音频语音转文字 · 完全本地 · 数据不出机
✨ 为什么选这个?
本项目 | xkiranj/ollama-vision-mcp | MikeyBeez/mcp-vision | |
图片识别 | ✅ | ✅ | ✅ |
视频分析 | ✅ 唯一! | ❌ | ❌ |
音频转录 | ✅ 唯一! | ❌ | ❌ |
中文模型 | ✅ MiniCPM 原生中文 | ❌ llava 英文为主 | ❌ |
安装 | 1 文件,丢进去就行 | Python venv + pip | macOS 限定 |
Windows | ✅ | ⚠️ | ❌ |
自动清理 | ✅ 分析完自动删视频 | ❌ | ❌ |
Related MCP server: api-vision-mcp
🚀 快速开始
1. 安装 Ollama 并拉取模型
ollama pull minicpm-v4.62. 安装依赖
npm install @ffmpeg-installer/ffmpeg @modelcontextprotocol/sdk
pip install faster-whisper3. 配置 AI Agent
Claude Desktop (claude_desktop_config.json):
{
"mcpServers": {
"vision": {
"command": "node",
"args": ["/path/to/vision_mcp_server.mjs"]
}
}
}Reasonix (reasonix.toml):
[[plugins]]
name = "vision"
command = "node"
args = ["/path/to/vision_mcp_server.mjs"]重启客户端即可使用。
🛠️ 工具
describe_image — 描述图片
输入: 图片文件的绝对路径
输出: 中文图片内容描述describe_video — 分析视频
输入: 视频文件路径 或 HTTP/HTTPS URL, interval (可选)
输出: 时间线帧描述 + AI 总结 + 自动清理describe_audio — 音频转文字 🆕
输入: 视频/音频文件路径 或 URL
输出: 音频元数据 + faster-whisper 语音转文字📊 视频分析流程
URL/文件 → 下载 → FFmpeg 提取关键帧 → MiniCPM-V 逐帧分析 → 汇总总结 → 自动清理🔧 依赖
FFmpeg via
@ffmpeg-installer/ffmpegfaster-whisper for audio transcription
🌍 适用场景
📸 让 DeepSeek 等纯文本模型看懂截图和照片
🎬 快速了解视频内容,不用逐秒观看
🎵 提取视频中的语音并转文字
📱 分析抖音/B站等平台视频
🔒 敏感图片/视频本地处理,数据不出机
🇨🇳 中文内容理解(MiniCPM 原生中文)
📄 许可证
MIT License
🙏 致谢
MiniCPM-V · Ollama · faster-whisper · MCP
This server cannot be deployed
Maintenance
Related MCP Connectors
Local ML: speech to text (SRT/VTT), text embeddings, reranking, similarity, entities, zero-shot.
Search photos and video moments by visual meaning, people, dates, and folders across local libraries, Apple Photos, and OneDrive. Manage collections, groups, favorites, people, and duplicate review. Synced metadata remains searchable while the desktop is offline; local and Apple Photos media access needs the owning device online with remote access enabled. OneDrive media can be accessed directly from the cloud. OAuth or Bearer API keys. Paid Cloud Access required; hosted visual search uses prepaid credits. Text-in-image search is semantic matching, not guaranteed exact OCR. Setup: https://organizewith.ai/en/guides/connect-ai-assistants
Image & PDF tools for AI agents: compress, convert, resize, PDF, AI vision, pipeline.
Image and video AI tools and your own pipelines, run from any AI assistant.
Related MCP Servers
- AlicenseBqualityDmaintenance一个MCP视觉分析工具,将本地图片交给豆包视觉模型分析,帮助Claude Code和DeepSeek理解截图内容。1ISC
- AlicenseNot gradedqualityDmaintenanceEnables image recognition using vision models via OpenAI-compatible APIs, supporting multiple platforms like OpenAI, DeepSeek, and Ollama.ISC
- AlicenseNot gradedqualityCmaintenanceProvides vision capabilities to text-only models (like DeepSeek) via Zhipu free vision models, enabling image analysis, OCR, and image comparison through natural language.MIT
- FlicenseNot gradedqualityAmaintenanceProvides offline, privacy-preserving image recognition, OCR, and scene description for AI assistants via the Model Context Protocol, with Vulkan-accelerated local processing.2-