Library

to-markdown

A versatile, TypeScript-first utility for turning files into Markdown. One promise-based call handles PDF, Word, HTML, spreadsheets, notebooks, feeds, archives, and media — with optional OCR for scanned pages and images.

• Package @cognipeer/to-markdown
• Licence MIT
• Runtime Node.js 18+

@cognipeer/to-markdown exists because retrieval pipelines, prompt contexts, and knowledge bases all want the same thing: clean Markdown, whatever the source file happened to be. Point it at a path, a Buffer, or a base64 string and it detects the format, picks the right converter, and hands back a Markdown string.

Introduction

Getting started

Install the package and run your first conversion from a path, a Buffer, or base64.

Guides

Format support

What each converter does, format by format, with example output.

Guides

OCR

Read text out of scanned PDFs and images with Tesseract or a vision model.

API Reference

API overview

Every exported function and type, with signatures and return values.

---

Install

npm

npm install @cognipeer/to-markdown

yarn

yarn add @cognipeer/to-markdown

pnpm

pnpm add @cognipeer/to-markdown

Quick example

import { convertToMarkdown } from '@cognipeer/to-markdown';

// Convert a PDF file
const markdown = await convertToMarkdown('./document.pdf');
console.log(markdown);

What it does

Detects the format for you

File extension, MIME type from a data URL, or magic bytes in a raw Buffer — with a plain-text fallback and a forceExtension escape hatch.

One converter per format

Each format has a specialised converter, so tables stay tables, notebook cells stay ordered, and feeds keep their structure.

TypeScript throughout

Written in TypeScript and shipped with full type definitions, as both ESM and CommonJS builds.

Supported formats

Format	Extensions	Description
PDF	.pdf	PDF documents (optional OCR for scans)
Word	.docx	Microsoft Word documents
HTML	.html, .htm	HTML web pages
Excel	.xlsx, .xls	Excel spreadsheets
CSV	.csv	Comma-separated values
Jupyter	.ipynb	Jupyter notebooks
PowerPoint	.pptx	PowerPoint presentations
XML/RSS	.xml, .rss, .atom	XML and feed formats
JSON	.json	JSON documents
YAML	.yaml, .yml	YAML documents
EPUB	.epub	EPUB e-books
Outlook	.msg	Outlook email messages
Images	.jpg, .jpeg, .png, .gif	Image files (metadata, optional OCR)
Audio	.mp3, .wav	Audio files (metadata extraction)
Text	.txt	Plain text files
Archives	.zip	ZIP archives

See Format Support for what each converter extracts.

Where to go next

• Getting Started — installation and the three input shapes.
• Converter Options — every option on ConverterOptions, with defaults.
• Format Support — format-by-format behaviour and sample output.
• Error Handling — failure modes, retries, and fallbacks.

About

Maintained by Cognipeer and released under the MIT License. Source on GitHub, published on npm.