How are equations handled?

Inline equations become $...$ and display equations become $$...$$ blocks. This format works in Obsidian, Notion, GitHub (with MathJax), and most Markdown renderers.

What about custom macros?

Custom LaTeX macros are expanded during conversion, so the output Markdown is self-contained and renders correctly without needing your macro definitions.

Are cross-references preserved?

Yes! References like \ref{sec:intro} become clickable links to the corresponding section, equation, or figure in the Markdown output.

Can I include figures?

Yes. Figures are included with their captions. You can optionally download assets as a ZIP file (Plus feature) to get the images alongside the Markdown.

Yes! You can convert papers programmatically via our API. Upload a paper, then fetch content in markdown format. See our API documentation for details.

LaTeXMarkdown

Free

LaTeX to Markdown Converter

Convert LaTeX papers to clean, readable Markdown with preserved equations and structure.

Free for LaTeX uploads

Convert Now API Docs

Free account required. See pricing for high-volume use.

Built for your workflow

LaTeX to Markdown conversion that actually works for academic papers.

Obsidian & Notion

Import papers into your knowledge base with working equations and links

LLM Context

Feed papers to ChatGPT, Claude, or your RAG pipeline in a clean format

GitHub README

Include paper content in documentation with proper math rendering

Blog Posts

Convert papers to blog-ready content with minimal editing

Semantic parsing

We understand LaTeX structure, not just text. That's why our output preserves what matters.

Semantic Parsing

Understands LaTeX structure - sections, equations, theorems, figures, tables as semantic elements

Cross-Reference Resolution

Automatically resolves \ref, \cite, and other cross-references to readable formats

Macro Expansion

Expands custom macros and commands so the output is self-contained

Bibliography Support

Includes formatted references with proper numbering and citation links

Equation Formatting

Inline equations as $...$ and display equations as $$...$$ - works everywhere

Section Anchors

Each section, equation, and figure gets a linkable ID for easy navigation

See what you get

Real output from converting the “Attention Is All You Need” paper.

output.md

# Attention Is All You Need

## Abstract

The dominant sequence transduction models are based on complex recurrent or convolutional neural networks...

## 1. Introduction

Recurrent neural networks, long short-term memory [#bib:1] and gated recurrent [#bib:2] neural networks...

The Transformer follows this overall architecture using stacked self-attention and point-wise, fully connected layers for both the encoder and decoder, shown in the left and right halves of Figure [#fig:1], respectively.

## 3.2. Attention

An attention function can be described as mapping a query and a set of key-value pairs to an output...

$$\text{Attention}(Q, K, V) = \text{softmax}\left(\frac{QK^T}{\sqrt{d_k}}\right)V$$ [#eq:1]

Equations, cross-references, and structure — all preserved.

How it works

Upload

Upload your .tex file or .zip archive with all source files

Process

We parse the LaTeX semantically — understanding sections, equations, and references

Download

Clean Markdown with $equations$, proper headings, and linkable references

Simple pricing