# TokenMarkdown.com
> Sub-150ms Web-to-Markdown Extraction Engine & Model Context Protocol (MCP) Server

## Overview
TokenMarkdown converts any public webpage into clean, deterministic GitHub-Flavored Markdown (GFM) optimized for LLM context windows, RAG pipelines, and autonomous AI coding agents.

## Core Endpoints
- POST https://tokenmarkdown.com/v1/extract : Structured web extraction returning clean markdown, title, author, and token estimates.
- POST https://tokenmarkdown.com/v1/extract/batch : Parallel extraction of up to 10 URLs in under 1 second.
- GET https://tokenmarkdown.com/https://<target-url> : Prefix proxy streaming clean raw markdown directly to stdout.
- GET https://tokenmarkdown.com/sse : Remote Model Context Protocol (MCP) event stream.

## Model Context Protocol (MCP) Tool
Run locally with zero configuration:
`npx -y tokenmarkdown-mcp`

Exposes tool: `extract_markdown` with parameters `url` (string, required), `include_images` (boolean), `include_links` (boolean), `selector` (string).

## Token Suite Network
- TokenEnrich.com : Sub-second B2B firmographics & DNS/tech stack qualification in ~180 tokens.
- TokenMarkdown.com : Sub-150ms web-to-markdown extraction engine.