CSV to Markdown table

How to convert CSV to a Markdown table

• Paste CSV or TSV into the box, or drop a .csv / .tsv file onto it. The delimiter (comma, semicolon, tab, or pipe) is detected — override it if a file uses something unusual.
• Tell it whether the first row is a header. If not, it generates Column 1, Column 2… (a Markdown table always needs a header).
• Pick pretty (padded, readable in a text editor) or compact output, and how columns align — numbers are right-aligned automatically.
• Copy the Markdown table or download it as a .md file. The rendered preview shows exactly how GitHub will display it.

The CSV parser follows RFC 4180, so the awkward parts of real files just work: fields wrapped in quotes that contain commas, doubled "" quotes, line breaks inside a field, a UTF-8 byte order mark from Excel, and rows with fewer cells than the header (padded, with a note telling you how many). Column widths are measured by display width, so Chinese, Japanese, and Korean text and emoji line up in the raw Markdown too.

Markdown table syntax, explained

A Markdown table (the GitHub-flavored Markdown extension that GitHub, GitLab, Obsidian, Notion imports, and most static site generators support) is three things: a header row, a delimiter row of dashes, and data rows, with cells separated by pipes.

| Item      | Region | Units |   Price |
| :-------- | :----: | ----: | ------: |
| Widget    |  EMEA  | 1,200 |   19.99 |
| Gadget    |  APAC  |    85 |  249.00 |
| A \| B   |   —    |     7 |    5.00 |

• Colons in the delimiter row set alignment: :--- left, :---: center, ---: right, plain --- default.
• The outer pipes are optional and the padding is cosmetic — renderers ignore it. The formatter adds it so the source stays readable.
• A literal pipe inside a cell is written \|; a line break is written <br>.
• Cells can hold inline Markdown — **bold**, `code`, links — but not block content like lists or multiple paragraphs.

Convert a Markdown table to CSV format

The Markdown → CSV tab reverses the process: paste a table copied from a README, a GitHub issue, or an LLM answer and get clean CSV. It finds every table in a longer document (skipping code blocks), so you can paste a whole .md file and choose the table you need. Escaped pipes are unescaped, <br> becomes a real line break inside a quoted field, and you can strip inline formatting so **42** becomes plain 42.

Opening the result in Excel? Use a semicolon delimiter if your Excel uses a comma as the decimal separator (most of europe), and tick "Add BOM" so accented characters and emoji survive the import. To go straight into a spreadsheet file, the Markdown to Excel converter downloads a real .xlsx with numbers typed as numbers, and the Markdown table to JSON converter turns the same table into an array of objects.

A Markdown table generator and formatter

Not starting from a CSV? The Table generator tab is a small spreadsheet that writes Markdown: type into cells, add rows and columns, set alignment per column, and copy the result. Press Enter to move down a row (a new row is added at the end), or paste a block of cells copied from Excel or Google Sheets and it fills the grid. Any table you convert or format can be opened in the grid with "Edit as grid".

The Markdown table formatter fixes tables that were typed by hand or generated without padding: it re-aligns every column to its widest cell, normalizes the delimiter row, keeps (or changes) each column's alignment, adds a missing delimiter row, and pads short rows — across every table in a document at once.

CSV to Markdown in Python, pandas, and the command line

If you convert tables as part of a script, these do the same job:

import pandas as pd

df = pd.read_csv("data.csv")
print(df.to_markdown(index=False))           # pip install tabulate
print(df.to_markdown(index=False, tablefmt="github"))

# pandoc 2.9+ reads csv directly
pandoc -f csv -t gfm data.csv -o table.md

# csvkit renders a markdown-compatible table
csvlook data.csv

Going the other way, pandas has no read_markdown; the usual trick is pd.read_csv(path, sep="|") followed by dropping the empty edge columns, stripping whitespace, and removing the delimiter row — which is exactly the fiddly part this page does for you, with escaped pipes handled correctly.

Frequently asked questions