AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
RAG / Knowledge Base/
markdownify-mcp
markdownify-mcp logo

markdownify-mcp

Active·★ 3.0k·MIT·Updated 2026-10-26
★ Essential★ Trending

Markdownify is an MCP server that converts many kinds of content into Markdown so that an AI model can read it. It handles PDF, images (with metadata), audio (with transcription), DOCX, XLSX and PPTX files, plus web content: YouTube video transcripts, Bing search results and general web pages. A separate `get-markdown-file` tool retrieves an existing .md or .markdown file, and the optional `MD_SHARE_DIR` environment variable restricts which directory it may read from.

Setup is from source: clone the repository, run `pnpm install` (which also installs `uv` and the related Python dependencies), `pnpm run build` and `pnpm start`, then point a desktop app's MCP config at `dist/index.js` with an absolute path. Tools are defined in `src/tools.ts` and the server in `src/server.ts`, so adding a converter means editing those files. The README opens with a request for a Windows user to help test Windows support, because pull requests exist that the maintainer cannot test. It is MIT licensed.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool. Data as of 2026-10-25.

License
MIT
Latest release
v1.1.0 · 2026-05-01
Last commit
2026-09-25
Contributors
9
Forks
255
Open issues and PRs
31
First published
2024-12-18

Source: GitHub, npm and PyPI public APIs.

#MCP Server#Markdown Conversion#Document Processing#Web Scraper#File Converter#Web Browsing
$ Install
$ pnpm install
↗ Visit site★ GitHub
01

Features

01Convert PDF, Images, Audio (with transcription), DOCX, XLSX, PPTX to Markdown
02Convert YouTube video transcripts, Bing search results, and general web pages to Markdown
03Retrieve existing Markdown files from a specified directory
04Model Context Protocol (MCP) server integration
05Support for `uv` and Python dependencies for underlying conversions
02

Why choose it

+One server covers PDF, Office files, images, audio and web pages
+Audio transcription and YouTube transcripts are included
+`MD_SHARE_DIR` lets you restrict which folder file retrieval can read
+Small codebase: tools live in a single file, easy to extend
03

Trade-offs

!No packaged release in the README: you clone, build and point your client at a local path
!Windows support is not confirmed; the maintainer asks for help testing it
!`pnpm install` also pulls in `uv` and Python dependencies, so both toolchains are needed
!The README does not say which libraries do the conversions or how well they handle complex layouts
05

Maintenance check

SlowingChecked: 2026-10-27
Last commit
2026-05-01
Commits, last 90 days
0
Releases, last 90 days
0
Latest release
v1.1.0 (2026-05-01)
Open issues
31
Recent issues with at least one comment
3/14
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #95 webpage-to-markdown fails on Apple Silicon (ARM) — onnxruntime CPU warning + crash5 comments · 2026-04-10
  • #148 Added 1-click config generator and setup guide for markdownify-mcp on MCP ServerList0 comments · 2026-10-20
  • #139 [audit N6] write_* family has empty default allowed_directories — discuss default-deny0 comments · 2026-08-29
  • #138 [audit N5] stderr from repomix is silently dropped (informational)0 comments · 2026-08-29
  • #137 [audit N4] inferExtensionFromUrl ignores ?query and #fragment0 comments · 2026-08-29

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

04

Who it fits

Pick it if
  • You want one simple server that turns PDF, Office files, images, audio and web pages into Markdown, including YouTube transcripts.
  • You want to extend it yourself: the tools live in one file, `src/tools.ts`.
  • You want to restrict file reads with `MD_SHARE_DIR`.
Look elsewhere if
  • You want a packaged install: you clone, build and point your client at a local path.
  • You are on Windows: support is not confirmed and the maintainer asks for testers.
  • Complex layouts matter: the README does not say which libraries do the conversions or how they handle them.
06

How it differs

anymd logovs anymd★ 1.1k

anymd is a Rust tool with local OCR, ASR, token budgets and citation anchors. Markdownify is simpler, and adds YouTube transcripts and Bing search results.

Full comparison →
07

Compatibility

Model Context Protocol (MCP)
Native
Verified via docs
Desktop Apps (MCP enabled)
Supported
Verified via docs
Node.js
Native
Verified via docs
Python Environment (via uv)
Supported
Verified via docs
Windows
Partial/Needs Help
Verified via docs
08

Quick start

1
$ pnpm install
09

Use cases

↳Preparing diverse content (documents, web pages) for AI model context
↳Archiving web content or multimedia in a standardized Markdown format
↳Automating conversion of various file types to a uniform text format
10

Alternatives

⚡Top VS:
VSanymd★1.1k→VSGPT Researcher★30.3k→VSatlas-mcp-server★478→VSMiroThinker★8.4k→
anymd logo
anymd★ 1.1k
Any file → clean Markdown for AI agents: PDF, Word, PowerPoint, Excel, EPUB, HTML and web pages, images (OCR), audio and video (metadata, subtitles, transcripts). A fast Rust MCP server and CLI that runs on your machine. No API key.
VSCompare→
GPT Researcher logo
GPT Researcher★ 30.3k
An autonomous agent that conducts deep research on any data using any LLM provider, producing detailed reports with citations from web and local sources.
VSCompare→
atlas-mcp-server logo
atlas-mcp-server★ 478
A Model Context Protocol (MCP) server for ATLAS, a Neo4j-powered task management system for LLM Agents - implementing a three-tier architecture (Projects, Tasks, Knowledge) to manage complex workflows. Now with Deep Research.
VSCompare→
MiroThinker logo
MiroThinker★ 8.4k
MiroThinker is an open source deep research agent optimized for research and prediction. It achieves a 80.8% Avg@8 score on the challenging GAIA benchmark.
VSCompare→
deep-research logo
deep-research★ 4.7k
Use any LLMs (Large Language Models) for Deep Research. Support SSE API and MCP server.
VSCompare→
ragflow logo
ragflow★ 93.0k
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
VSCompare→
ICLR2025-Papers-with-Code logo
ICLR2025-Papers-with-Code★ 589
历年ICLR论文和开源项目合集，包含ICLR2021、ICLR2022、ICLR2023、ICLR2024、ICLR2025.
VSCompare→
obsidian-mcp-tools logo
obsidian-mcp-tools★ 827
Add Obsidian integrations like semantic search and custom Templater prompts to Claude or any MCP client.
VSCompare→
See all alternatives →

Featured in these lists

  • Best Tools to Turn PDFs and Office Files into Markdown for AI Agents

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs05Maintenance check04Who it fits06How it differs07Compatibility08Quick start09Use cases10Alternatives
zcaceres /
markdownify-mcp
↗

MCP server that converts files and web content to Markdown

3.0k255MIT
Last commit2d ago
StatusActive
CategoryRAG / Knowledge Base
Contributors9
Latest releasev1.1.0
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.