Manual setup required. The maintainer's config contains paths only you know - edit the placeholders below before adding it to Claude Code.
1
Prepare the server locally
Run this once before adding it to Claude Code.
pip install -r requirements.txt2
Register it in Claude Code
claude mcp add -e "VOLCENGINE_API_KEY=${VOLCENGINE_API_KEY}" doubao-image -- python /path/to/server.pyReplace any placeholder paths in the command with the real path on your machine.
Required:
VOLCENGINE_API_KEY3
Make your agent remember this setup
doubao-image's config, env vars, and the gotchas you hit — recalled in every future Claude Code, Cursor, and Codex session.
npx conare@latestFree · one command · indexes the sessions already on disk. Set up in the browser instead →
What it does
- Supports 18 image formats including JPG, PNG, GIF, WEBP, and HEIC
- Automatic image compression for large files
- Caching mechanism for repeated queries with sub-0.1s response time
- Optimized for Chinese language understanding
- Fast processing with 20-30 second average response time
Tools 1
describe_imageAnalyzes an image from a local path or URL and provides a descriptive text response.Environment Variables
VOLCENGINE_API_KEYrequiredAPI Key for accessing the Volcengine Doubao vision modelTry it
→Describe the contents of this image: https://example.com/photo.jpg
→Can you analyze the local image at D:/download/photo.jpg and tell me what is in it?
→Describe this image in English: photo.jpg
→What is the main subject of this image?
Original README from 9527wow/doubao-image-mcp
豆包图片描述 MCP 服务器
让 iFlow CLI 和 Claude Desktop 支持豆包视觉理解
🚀 快速开始（3 步）
1. 安装依赖
pip install -r requirements.txt
2. 配置 API Key
方式 1：环境变量（推荐）
# Windows
set VOLCENGINE_API_KEY=你的豆包API Key
# Linux/Mac
export VOLCENGINE_API_KEY=你的API Key
方式 2：配置文件
cp config.example.json config.json
# 编辑 config.json，填入你的 API Key
3. 添加到 iFlow CLI
iflow mcp add doubao-image-mcp python C:\Users\15579\doubao-image-mcp\server.py
4. 测试
iflow
> 描述: https://picsum.photos/800/600
📖 完整文档
- INSTALL_GUIDE.md - 详细安装和配置指南
- 论坛帖子.md - 完整的开发历程和技术细节
✨ 特性
- ⚡ 快速：20-30 秒响应
- 💾 缓存：重复查询 < 0.1 秒
- 🖼️ 18 种格式：JPG、PNG、GIF、WEBP、HEIC 等
- 🗜️ 自动优化：大图片自动压缩
- 🌏 中文优化：专为豆包模型优化
📂 项目包含
server.py- MCP 服务器主程序config.example.json- 配置示例requirements.txt- Python 依赖compress.py- 图片压缩工具diagnose_speed.py- 速度诊断工具- 完整文档
🆘 获取 API Key
- 访问 火山引擎控制台
- 开通"火山方舟"服务
- 创建 API Key
- 确保有视觉理解模型权限
💡 使用示例
# 描述本地图片
描述: D:/download/photo.jpg
# 描述网络图片
描述: https://example.com/image.jpg
# 自定义提示词
描述: photo.jpg，请用英文描述
📄 License
MIT License - 自由使用、修改、分发
🔗 相关链接
版本: 1.0.0
更新: 2026-02-03
分享者: [Your Name]