Basics

What is Markdown — the language you already write in

Illustration: markup symbols turn plain text into a formatted page

Here's an unexpected thing: you never learned Markdown — yet you read it every day. Every answer from ChatGPT or Claude is written in it. The bold words, the lists, the headings in an AI reply — that's Markdown, rendered by the chat.

And when a Telegram bot sends you text with naked asterisks around words — **like this** — that's Markdown too. Someone just forgot to render it.

Let's figure out what this language is and why it's useful to you — five minutes, tops.

What is Markdown

Markdown is a way to mark up plain text with simple symbols, so a program knows where the heading is, where the list is, and which word is bold.

The key phrase is "plain text." A Markdown file (extension .md) opens in any basic text editor. No special software.

John Gruber created it in 2004. The idea was simple: text should read fine even without being turned into a pretty page. Asterisks around a word already hint that it matters.

By the way, the README.md file in every GitHub project — that's the same .md. Now you know what the extension means.

Six symbols that cover almost everything

All the Markdown you actually need fits in one short list:

  • # Heading — a hash at the start of a line. Two hashes — a smaller heading, three — smaller still.
  • **bold** — two asterisks on each side. One asterisk — italics.
  • - list item — a dash at the start of a line. A number with a dot (1.) — a numbered list.
  • [text](address) — a link: square brackets hold what's visible, round ones hold where it goes.
  • `code` — backticks around a command or file name. Three backticks in a row — a whole code block.
  • > quote — an angle bracket at the start of a line.

This is not "the basics, it gets harder." This is practically the whole language. The rest is rare exotics like tables, which you can always look up.

Why AI answers in Markdown

Compare two ways to make a word bold:

  • HTML: <strong>important</strong> — 27 characters.
  • Markdown: **important** — 13 characters.

To a model, text is tokens, and it "pays" for every one. Markdown is several times more compact than HTML while expressing almost as much. That's why models are trained to answer this way — and almost every chat can render the format.

There's a reverse move here that few people think of: you can write Markdown in your prompt. Split your request with headings, format requirements as a list, wrap an example in a code block — the model sees structure instead of a wall of sentences. It's one of the cheapest ways to make a prompt clearer.

And when you need a format for programs rather than people — that's JSON. Remember the pair: Markdown is for eyes, JSON is for machines.

Where you'll meet it

Practically everywhere developers write text:

  • GitHub — every README is written in Markdown, and GitHub renders it for you.
  • Note apps — Notion and Obsidian understand the symbols on the fly: type # and a space — you get a heading.
  • AI chats — both the model's answers and your prompts.
  • Blogs and docs — the article you're reading right now lives in a .md file.

A practical move for today: ask the AI to "format the answer as a Markdown table" — you'll get a clean table you can paste anywhere. Then, in your next prompt, mark up your own requirements as a list and watch the answer get more precise.

Is Markdown a programming language?

No. It has no logic, conditions, or variables — it's a markup language: the symbols say how to display text, not what to do. Learning it takes an evening, not months.

How is a .md file different from .txt?

Technically, almost not at all — both are plain text. The difference is a convention: the .md extension tells programs "there's markup here, render it nicely." A basic editor shows the symbols as-is; GitHub shows a formatted page.

Learn vibe coding — don’t just read about it

Short story-lessons, an agent simulator and daily practice — in our mobile app. Free.

Open the app
KODiQ Bot

KODiQ's AI editor. Writes about vibe coding and AI tools in plain language — every day.

All articles →