A
licenseA
qualityD
maintenanceEnables AI agents to analyze images, extract text, compare images, and analyze video through any OpenAI-compatible vision model.
4
151 npm
20
MIT
vision-eye
by ergou-yu
Reuses local Ollama vision models (e.g., Qwen-VL, MiniCPM-V, LLaVA) to provide image description, OCR, table parsing, UI analysis, and region zooming entirely on-device without cloud APIs.
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-eyeextract the text from this 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-eye
纯本地视觉插件包：给纯文本大模型（DeepSeek、Qwen 纯文本版等）接上"眼睛"。
MCP server（工具层）+ Claude Agent Skill（触发层）+ 本地视觉模型（推理层）。
零云端 API、零 API key，图片不出本机（仅首次模型下载需联网，支持
HF_ENDPOINT=https://hf-mirror.com 镜像）。
安装
pip install vision-eye # 形态 A：复用本机 Ollama 视觉模型
vision-eye install # 自动检测环境并分级
vision-eye doctor # 诊断后端 / 模型 / 降级历史安装形态
形态 | 条件 | 行为 |
A | Ollama 已有视觉模型（qwen*vl / minicpm-v / llava 等，模糊匹配） | 复用，不下载 |
B-GPU | CUDA 可用且显存 ≥ 8GB | Qwen2.5-VL-3B int4 + SmolVLM-500M 兜底 |
B-CPU | 无 GPU / 显存或磁盘不足 | 仅 SmolVLM-500M |
Related MCP server: Vision Bridge MCP Server
MCP 注册（Claude Code）
claude mcp add vision-eye -- vision-eye serve工具
工具 | 功能 |
describe_image | 通用描述（focus 可选） |
extract_text | OCR，保留布局 |
parse_table | 表格转 markdown |
analyze_ui | 截图元素识别与报错诊断 |
zoom_region | 放大局部细看 (x, y, w, h) |
返回 < 500 token，超长截断并提示用 zoom_region
压缩到最长边 2048px；MD5(压缩后字节+prompt) 缓存于
~/.vision-eye/cache/三级降级链自动兜底，降级结果带
[precision: low]标记（复杂图表结论需谨慎）降级日志：
~/.vision-eye/logs/，vision-eye doctor查看
CLI
vision-eye install / doctor / serve / daemon / cache cleardaemon 常驻模式预加载模型，提供 streamable HTTP（默认 127.0.0.1:8765）。
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo issues
Related MCP Connectors
Turn any LLM multimodal; generate images, voices, videos, 3D models, music, and more.
E2LLM gives your AI eyes and hands in a real browser: structured perception (SiFR) plus action.
Image & PDF tools for AI agents: compress, convert, resize, PDF, AI vision, pipeline.
Give agents eyes on any web page: structured context, and changes explained in plain language.
Related MCP Servers
- AlicenseNot gradedqualityBmaintenanceEnables text-only LLMs to understand images by converting them into text descriptions, supporting multiple vision backends like cloud APIs, local models, and OCR engines.1MIT
- AlicenseNot gradedqualityAmaintenanceProvides local image understanding for text-only LLMs with tools for image analysis, OCR, object detection, and cropping, all processed on-device.2MIT
- AlicenseNot gradedqualityBmaintenanceEnables text-only LLMs to analyze, OCR, and compare images by forwarding them to vision-capable models via OpenAI or Anthropic APIs.28 npmMIT