Your LLM Writes $x$, Your Markdown Parser Wants $x$

TL;DR — LLMs love $...$ and \[...\]. Most Markdown math plugins only speak $...$ and $$...$$. I tried to bridge that with a regex, failed in interesting ways, and ended up teaching the tokenizer instead. Two packages came out of it: micromark-extension-math-extended and remark-math-extended.

The setup

Sometimes the hard part of rendering math isn't the equation. It's agreeing on where the equation starts and stops.

I was piping Markdown from some OpenAI models into a rendering pipeline, and the output kept looking like this:

The lift coefficient is $C_L$.

Your LLM Writes \(x\), Your Markdown Parser Wants $x$

Related reading

How I Stopped Fighting Regex and Finally Extracted Data with LLMs

I Fixed LLM Markdown Errors with Jinja2 and AST Parsing

Reduce LLM Token Waste in RAG with Markdown

MarkItDown: Microsoft's Tool for Converting Almost Anything to Markdown

Why My Regex-Based Parser Failed and How LLM Function Calling Saved Me

Stop letting your AI guess: The case for deterministic Regex in MCP