About LLMtoMD
LLMtoMD is a hosted MCP server for converting documents, web pages, and audio into Markdown, then searching and asking questions across saved material. It is for people who want an AI assistant to work with converted content and a document library.
Key features
- 01Convert documents, web pages, and audio
- 02Handle scanned PDFs with OCR
- 03Store converted documents in a library
- 04Semantic search and source-backed answers
- 05Record and check coding-task file scope
Use cases
- 01Convert a PDF or web page for AI use
- 02Search across a collection of converted files
- 03Ask questions and review source documents
- 04Track file boundaries during coding tasks
MCP support
Provides conversion, document-library, search, question-answering, and coding-scope tools to MCP clients through a hosted Streamable HTTP endpoint. Authentication uses OAuth 2.1.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://mcp.llmtomd.com/mcpclaude mcp add --transport http com-llmtomd-llmtomd https://mcp.llmtomd.com/mcp{"mcpServers":{"com-llmtomd-llmtomd":{"type":"http","url":"https://mcp.llmtomd.com/mcp"}}}Open in Cursor{"servers":{"com-llmtomd-llmtomd":{"type":"http","url":"https://mcp.llmtomd.com/mcp"}}}Open in VS Code{"mcpServers":{"com-llmtomd-llmtomd":{"type":"http","url":"https://mcp.llmtomd.com/mcp"}}}{"mcpServers":{"com-llmtomd-llmtomd":{"serverUrl":"https://mcp.llmtomd.com/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMELLMtoMD MCP Server
Turn PDFs, Office documents, websites and audio into clean, AI-ready Markdown, and let your AI assistant search and answer questions across everything you've converted.
LLMtoMD is a hosted MCP server. There's nothing to install or run: connect your client to the endpoint below and sign in with your LLMtoMD account.
| Endpoint | https://mcp.llmtomd.com/mcp |
| Transport | Streamable HTTP |
| Auth | OAuth 2.1 (PKCE + dynamic client registration). Your client opens a sign-in page; no API keys to copy. |
| Registry | com.llmtomd/llmtomd on the official MCP Registry |
| Account | Free sign-up at llmtomd.com |
Connect
VS Code (GitHub Copilot)
Or add it by hand: Command Palette → MCP: Add Server → HTTP, enter https://mcp.llmtomd.com/mcp, and sign in when prompted. You can also add it to .vscode/mcp.json:
{
"servers": {
"llmtomd": { "type": "http", "url": "https://mcp.llmtomd.com/mcp" }
}
}
Claude (web and desktop)
Settings → Connectors → Add custom connector, paste https://mcp.llmtomd.com/mcp, then sign in.
ChatGPT
Settings → Apps & Connectors → Create (developer mode), use https://mcp.llmtomd.com/mcp with OAuth, then sign in.
Cursor and other clients
Any MCP client that supports remote Streamable HTTP servers with OAuth works. Point it at https://mcp.llmtomd.com/mcp.
Tools
Convert
| Tool | What it does |
|---|---|
convert_url | Convert a web page or a publicly reachable file URL into Markdown |
convert_file | Convert a small local file (sent inline) into Markdown |
upload_file | Get a one-time upload URL for large or binary files (big PDFs, scans, audio, video) |
convert_upload | Convert a file uploaded with upload_file |
Your library
| Tool | What it does |
|---|---|
list_documents | List documents in your LLMtoMD library, newest first |
get_document | Fetch a stored document's full Markdown |
search_documents | Semantic search across your documents |
ask_documents | Answer a question from your documents, with sources |
list_collections | List your collections (projects) and their document counts |
save_note | Save text or Markdown into your library so it's searchable later |
Coding-task scope guard
| Tool | What it does |
|---|---|
start_task | Record the approved file boundary of a coding task before editing |
check_scope | Check changed files against that boundary |
approve_scope | Widen the boundary when a human approves extra files |
close_task | Mark a task finished |
list_tasks | List task scopes and their status |
Supported formats
PDF (including scanned PDFs via OCR), Word (DOCX), PowerPoint (PPTX), Excel (XLSX), HTML and whole-site crawls, images, and audio/video transcription. See llmtomd.com/formats.
Privacy and security
Source files are processed and deleted automatically within 24 hours. Converted documents stay in your account until you delete them. You can disconnect a client at any time from your LLMtoMD dashboard, which cuts its access immediately. See llmtomd.com/security and llmtomd.com/privacy.
Pricing
A free plan is available, and conversions use credits. See llmtomd.com/pricing.
Support
Email [email protected] or open an issue in this repository.
This repository documents the hosted LLMtoMD MCP server. The service itself is closed source; the documentation here is MIT licensed.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.