Markitdown MCP Server Cba7c082

MCP开发工具

提供方 RECTOR-LABS

开发工具 未标注 Token 社区 / 第三方 Claude DesktopClaudeCursor

Markitdown MCP Server是一款云端服务,能够将多种格式的文档快速转换为AI优化的Markdown格式,适用于RAG管道、知识库迁移和研究分析等场景。

一句话选型结论

按资料字段整理

Markitdown MCP Server Cba7c082 属于开发工具类 MCP 服务器,传输方式未标注,Token,可直接接入 Claude Desktop、Claude、Cursor,没有现成安装命令,需按官方文档接入。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类开发工具 传输方式未标注 鉴权Token 部署方式未标注
GitHub 星标 1 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

开发工具
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

Shell本地部署文档转换Markdown处理云端服务AI优化多格式支持

安装与配置

未标注 Token
上游未给出传输方式
  1. 1
    准备运行环境

    按官方文档确认运行环境(常见为 Node.js 或 Python),装好后再执行安装命令。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
代码与仓库操作

在对话里直接查代码、提 issue、走 PR,省掉来回切工具。

本地命令与脚本

把构建、测试、格式化这类重复命令交给模型执行。

接口联调与测试

让模型调你的接口、跑一遍用例,快速定位问题。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 已适配 Claude Desktop、Claude、Cursor,换客户端不用重新研究配置。

先想清楚的地方

  • 没有现成安装命令,需要按官方文档自己填启动命令和参数。
  • 使用前要准备好Token对应的凭据,并按最小权限配置。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在开发工具场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
Markitdown MCP Server Cba7c082当前 未标注 未标注 未标注 无 2026-05-17
Git 838e2ec1 12 stdio · 本地进程 本地部署 有 2026-05-17
Markitdown MCP 2c653a5f 1 stdio · 本地进程 本地部署 有 2026-05-17
JavaGuide 未标注 stdio · 本地进程 本地部署 有 2026-04-11
awesome-claude-skills 未标注 stdio · 本地进程 本地部署 有 2026-04-26
Context7 831acf40 未标注 stdio · 本地进程 本地部署 有 2026-05-17

服务器档案

开发工具
字段来自项目仓库与公开索引,中文说明由本站标注
提供方RECTOR-LABS上游标注的维护方 RECTOR-LABS
分类开发工具本站按上游分类展示
连接方式未标注上游未给出传输方式
鉴权方式Token远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · 未标注本地安装时用到的包标识与运行时
部署类型未标注上游给出的部署形态
最近核验2026-05-17上游最后核验时间

详细介绍

12.0 KB
来自项目仓库与上游目录的原文,未做删改

Markitdown MCP Server ⚡

> Convert any document to AI-ready Markdown in seconds
> Cloud-hosted Model Context Protocol server powered by Microsoft's Markitdown


🎯 What is This?

Markitdown MCP Server is a cloud-hosted service that converts documents into clean, AI-optimized Markdown. Built on Microsoft's Markitdown library (82k+ ⭐), it eliminates the need for local Python installations and provides instant, scalable document conversion through the Model Context Protocol.

Perfect for RAG pipelines, knowledge bases, AI agents, and document processing workflows.


✨ Key Features

🚀 Universal Format Support

Convert 29+ file formats to clean Markdown:
- Documents: PDF, DOCX, PPTX, XLSX
- Images: PNG, JPG, GIF (with OCR)
- Web: HTML, XML
- Audio: MP3, WAV (with transcription)
- Archives: ZIP (extract and convert contents)
- And many more!

☁️ Zero Setup Required

  • No Python installation needed
  • No dependency management
  • No local configuration
  • Just call the API and get Markdown

🎭 MCP Native

  • First-class Model Context Protocol support
  • Works seamlessly with Claude Desktop, Cursor, Aider
  • AI agents can discover and use it automatically

⚡ Lightning Fast

  • Direct Python library integration (no subprocess overhead)
  • Typical conversion: < 3 seconds
  • Cloud-scale infrastructure via Apify

💰 Pay-Per-Use

  • $0.01 per Actor start
  • $0.02 per document conversion
  • No subscriptions, no minimums

🎬 Quick Start

For AI Users (Claude Desktop)

  1. Add to MCP Configuration

Create or edit ~/Library/Application Support/Claude/claude_desktop_config.json:```json
{
"mcpServers": {
"markitdown": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://rector-labs--markitdown-mcp-server.apify.actor/mcp",
"--header",
"Authorization: Bearer YOUR_APIFY_TOKEN"
]
}
}
}


2. **Restart Claude Desktop**

3. **Convert Documents**

Simply ask Claude:```
&quot;Convert this PDF to markdown: https://example.com/document.pdf&quot;

Claude will automatically use the Markitdown tool!


For Developers (API)

Direct HTTP Request```bash

curl -X POST https://api.apify.com/v2/acts/rector_labs~markitdown-mcp-server/runs \
-H "Authorization: Bearer YOUR_API_TOKEN" \
-H "Content-Type: application/json" \
-d '{
"fileUrl": "https://example.com/document.pdf"
}'


#### Python Example```python
from apify_client import ApifyClient

client = ApifyClient(&#x27;YOUR_API_TOKEN&#x27;)
run = client.actor(&#x27;rector_labs/markitdown-mcp-server&#x27;).call(
    run_input={
        &#x27;fileUrl&#x27;: &#x27;https://example.com/document.pdf&#x27;
    }
)

# Get markdown output
for item in client.dataset(run[&#x27;defaultDatasetId&#x27;]).iterate_items():
    print(item[&#x27;markdown&#x27;])

JavaScript/TypeScript Example```typescript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_API_TOKEN' });

const run = await client.actor('rector_labs/markitdown-mcp-server').call({
fileUrl: 'https://example.com/document.pdf'
});

// Get markdown output
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items[0].markdown);


---

## 📚 Supported Formats

### Documents &amp; Spreadsheets

| Format | Extension | Notes |
|--------|-----------|-------|
| PDF | `.pdf` | Text extraction, OCR support |
| Word | `.docx`, `.doc` | Preserves formatting |
| PowerPoint | `.pptx`, `.ppt` | Slide text extraction |
| Excel | `.xlsx`, `.xls` | Table to Markdown |
| CSV | `.csv` | Table formatting |
| TSV | `.tsv` | Table formatting |

### Images

| Format | Extension | Notes |
|--------|-----------|-------|
| PNG | `.png` | OCR text extraction |
| JPEG | `.jpg`, `.jpeg` | OCR text extraction |
| GIF | `.gif` | OCR text extraction |
| BMP | `.bmp` | OCR text extraction |

### Web &amp; Markup

| Format | Extension | Notes |
|--------|-----------|-------|
| HTML | `.html`, `.htm` | Clean conversion |
| XML | `.xml` | Structured data |
| Markdown | `.md` | Pass-through |

### Audio &amp; Video

| Format | Extension | Notes |
|--------|-----------|-------|
| MP3 | `.mp3` | Speech-to-text transcription |
| WAV | `.wav` | Speech-to-text transcription |
| YouTube | URLs | Transcript extraction |

### Archives

| Format | Extension | Notes |
|--------|-----------|-------|
| ZIP | `.zip` | Extract and convert contents |

---

## 💡 Use Cases

### 🤖 RAG Pipelines```
PDF Documents → Markitdown → Clean Markdown → Vector DB → LLM

Perfect for preparing documents for semantic search and retrieval.

📖 Knowledge Base Migration

Convert legacy documentation (PDFs, Word docs) to modern Markdown format for wikis, documentation sites, or content management systems.

🎓 Research & Academia

Extract text from research papers, presentations, and datasets for analysis and processing.

📊 Data Extraction

Convert invoices, reports, and spreadsheets into structured Markdown for further processing.

🔄 Batch Processing

Process hundreds of documents in parallel using Apify's infrastructure.


🔌 Integrations

Claude Desktop```json

{
"mcpServers": {
"markitdown": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://rector-labs--markitdown-mcp-server.apify.actor/mcp",
"--header",
"Authorization: Bearer YOUR_APIFY_TOKEN"
]
}
}
}


### n8n Workflow
1. Add **Apify** node
2. Select **Markitdown MCP Server** actor
3. Configure file URL input
4. Connect to downstream nodes

### Make.com (Integromat)
1. Add **Apify** module
2. Select actor: `rector_labs/markitdown-mcp-server`
3. Map file URL from trigger
4. Use output in next steps

### Zapier
1. Choose **Apify** app
2. Action: **Run Actor**
3. Actor: `markitdown-mcp-server`
4. Map data from previous steps

---

## ⚙️ Configuration

### Input Parameters

| Parameter | Type | Required | Description |
|-----------|------|----------|-------------|
| `fileUrl` | string | ✅ (or base64) | URL of the document to convert |
| `fileBase64` | string | ✅ (or URL) | Base64-encoded file content |

**Note**: Provide **either** `fileUrl` **or** `fileBase64`, not both.

### Example Inputs

**URL-based**:```json
{
  &quot;fileUrl&quot;: &quot;https://example.com/document.pdf&quot;
}

Base64-based:```json
{
"fileBase64": "JVBERi0xLjQKJeLjz9MKMyAwIG9iago8PC..."
}


---

## 📊 Output Format

The actor outputs clean Markdown text with metadata:```json
{
  &quot;event&quot;: &quot;conversion_success&quot;,
  &quot;file_size&quot;: 153600,
  &quot;markdown_length&quot;: 5234,
  &quot;file_type&quot;: &quot;.pdf&quot;
}

The Markdown content is returned as the tool response.


💲 Pricing

Pay-Per-Event Model

Event Price Description
Actor Start $0.01 One-time fee per Actor run
Document Conversion $0.02 Per successful conversion

Example Costs

  • Single document: $0.03 total ($0.01 start + $0.02 conversion)
  • 100 documents: ~$2.10 ($0.01 start + $2.00 conversions)
  • 1,000 documents: ~$20.10 ($0.01 start + $20.00 conversions)

No subscriptions. No minimums. Pay only for what you use.


🚀 Performance

Metric Value
Average conversion time < 3 seconds
Small files (< 1MB) < 2 seconds
Large files (10MB+) < 10 seconds
Concurrent processing Unlimited (cloud-scaled)
Uptime 99.95% (Apify SLA)

🛠️ Advanced Features

Error Handling

The actor gracefully handles:
- Invalid file URLs (404, network errors)
- Unsupported file formats (clear error messages)
- Corrupted files (validation before processing)
- Large files (automatic timeout handling)

Logging & Debugging

All conversions are logged with:
- File type and size
- Conversion duration
- Success/failure status
- Error details (if any)

Custom Options

Coming soon:
- Azure Document Intelligence integration
- OpenAI image description
- Custom OCR settings
- Batch processing mode


🔒 Security & Privacy

  • No data retention: Files are processed and immediately deleted
  • Encrypted transport: All transfers use HTTPS
  • Isolated execution: Each conversion runs in a sandboxed container
  • No logging of content: Only metadata is logged
  • GDPR compliant: Hosted on Apify's secure infrastructure

❓ FAQ

Q: What's the difference between this and running Markitdown locally?

A: This is a cloud-hosted service with:
- ✅ No Python installation required
- ✅ No dependency management
- ✅ Automatic scaling for batch processing
- ✅ MCP integration for AI agents
- ✅ 99.95% uptime guarantee
- ✅ Pay-per-use (no server costs)

Q: Can I convert password-protected PDFs?

A: Not currently. Password-protected documents will return an error. Remove protection before conversion.

Q: What's the maximum file size?

A: 100 MB hard limit. Files over 50 MB may take longer to process. For larger files, consider splitting them first.

Q: Does it work with scanned PDFs (images)?

A: Yes! OCR (Optical Character Recognition) is supported for image-based PDFs and image files.

Q: Can I use this in production?

A: Absolutely! The actor runs on Apify's production infrastructure with 99.95% uptime SLA.

Q: How accurate is the Markdown output?

A: Markitdown preserves:
- ✅ Headings and structure
- ✅ Bold and italic formatting
- ✅ Lists (ordered and unordered)
- ✅ Tables
- ✅ Links
- ✅ Code blocks

Complex layouts may need manual review.

Q: Can I convert multiple files at once?

A: Yes! Run multiple Actor instances in parallel, or use batch mode (contact for enterprise pricing).


🐛 Troubleshooting

"File download failed: HTTP 404"

Cause: The URL is invalid or the file doesn't exist.

Solution:
- Verify the URL is correct and publicly accessible
- Ensure the file hasn't been deleted or moved
- Check for authentication requirements

"Unsupported file format"

Cause: The file extension is not in the supported formats list.

Solution:
- Check the Supported Formats section
- Convert the file to a supported format first
- Contact support if you need a specific format added

"Conversion timeout"

Cause: The file is too large or complex.

Solution:
- Split large files into smaller chunks
- Simplify complex documents
- Increase timeout (contact support for enterprise plans)

"Invalid base64 content"

Cause: The base64 string is malformed or incomplete.

Solution:
- Verify base64 encoding is correct
- Ensure no truncation occurred during transfer
- Use fileUrl instead if possible


📖 Documentation

  • MCP Protocol: modelcontextprotocol.io
  • Microsoft Markitdown: github.com/microsoft/markitdown
  • Apify Platform: docs.apify.com
  • Python SDK: docs.apify.com/sdk/python

🤝 Support

Need Help?

  • 📧 Email: support@apify.com
  • 💬 Discord: apify.com/discord
  • 📚 Documentation: docs.apify.com
  • 🐛 Bug Reports: GitHub Issues

Community

  • ⭐ Star on GitHub: RECTOR-LABS/markitdown-mcp-server
  • 🐦 Follow Updates: @apify
  • 💡 Feature Requests: Open a GitHub issue

🚀 Get Started Now

Deploy to Apify

  1. Log in to Apify```bash
    apify login

2. **Deploy the Actor**```bash
apify push
  1. Enable Standby Mode

Go to Actor settings and enable standby mode.

  1. Get Your Actor URL

Your MCP endpoint will be: https://rector-labs--markitdown-mcp-server.apify.actor/mcp

  1. Connect AI Agents

Add the endpoint to Claude Desktop, Cursor, or your favorite MCP client!


📜 License

This project is built on:
- Microsoft Markitdown: MIT License
- Apify SDK: Apache 2.0 License
- MCP SDK: MIT License

Actor code: MIT License


🙏 Credits

Built with:
- Microsoft Markitdown - Document conversion library (82k+ ⭐)
- Apify Platform - Serverless cloud infrastructure
- MCP Protocol - AI agent integration standard


Made with ❤️ for the AI developer community

Powered by Apify •
MCP Protocol •
Microsoft Markitdown

常见问题

6 问
答案由本站依据条目字段整理,不引用上游原文
Markitdown MCP Server Cba7c082 怎么安装?

资料没有给出安装命令,请到官方仓库或文档按说明接入;这类条目多数是生态应用,或尚未发布可安装包。

Markitdown MCP Server Cba7c082 怎么在 Claude Desktop、Claude、Cursor 里配置?

它的传输方式未标注,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

Markitdown MCP Server Cba7c082 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

Markitdown MCP Server Cba7c082 需要 API Key 吗?

资料标注需要Token认证,使用前准备好对应凭据,并按最小权限配置。

Markitdown MCP Server Cba7c082 支持哪些客户端?

资料标注支持 Claude Desktop、Claude、Cursor;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

Markitdown MCP Server Cba7c082 能提供哪些能力?

资料显示能力标签包括 多格式支持、文档转换、Shell、Claude。具体工具清单以官方文档为准。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-05-17。