Type your table, or paste it in from Excel, Google Sheets, a CSV file or an existing Markdown table. The Markdown is written underneath as you go, and it goes back into a spreadsheet just as easily.
Click a cell and paste. A table copied from Excel, Google Sheets, a web page or a chatbot fills in from that cell. The first row is the header.
Only the text above changes. Both versions look exactly the same once rendered.
Two ways to get the same table into a spreadsheet.
Copies the table so that every cell lands in its own column when you paste it into a sheet. Bold, links and code marks are left out.
A file you can open in Excel, Numbers, LibreOffice or anything else that reads spreadsheets.
Most tables already exist somewhere before they need to be Markdown. They sit in a spreadsheet, a CSV export, a web page or an answer from a chatbot, and typing them out again by hand is slow and easy to get wrong.
So the editor takes a paste from any of those. Cells copied from Excel or Google Sheets land in the same rows and columns. A CSV is read with its quotes and separators handled properly. A Markdown table is read back into cells, alignment included, so you can fix a typo or add a column without counting pipes.
A Markdown table is read twice: once rendered on GitHub, in your docs or in a chat, and once as plain text by whoever edits the file next. With the columns lined up, the raw version is almost as easy to read as the rendered one.
The small details are taken care of. A pipe inside a cell is escaped so it does not split the column, a line break becomes <br>, and the alignment you set for each column is written into the row of dashes under the header.
It works in the other direction too. Paste a Markdown table into the editor and press Copy cells. Each value lands in its own column instead of everything piling up in one, and the bold marks and links are left behind.
The CSV download is saved in a way Excel reads correctly, so letters like æ, ü and é survive. Everything happens in your browser, and nothing you paste is sent anywhere.
Select the cells in Excel, copy them, click the first cell of the table here and paste. The rows and columns fill in as they were, and the Markdown appears underneath, ready to copy. Google Sheets, Numbers and LibreOffice work the same way. If what you have is a CSV file, use Open a CSV file instead.
Because the copy button in ChatGPT, Claude and Gemini gives you the table as Markdown: one line of text per row, with pipes between the cells. Excel does not read Markdown, so each line lands in a single cell. Paste the table into the editor here, press Copy cells, and paste again in your spreadsheet. The cells go into their own columns, and the bold marks and links are left behind.
As a Markdown table. Pasted spreadsheet cells often reach the chat as values separated by tabs, and it is hard to see from the prompt which number belongs to which column. A Markdown table keeps the header and the columns clear, both to the model and to you when you read the prompt back. Paste your cells here and copy the result into your message.
It depends on your Excel, not on the data. Where the comma is the decimal mark, as in most of Europe, Excel expects a semicolon between columns and opens a comma-separated file in a single column. If you are in the US, the UK or anywhere else that writes 1.5 with a point, the comma is right. When in doubt, try one and look at the result.
Yes. The file is saved as UTF-8 with a small marker at the start that tells Excel which encoding it is. Without that marker Excel guesses, usually wrongly, and letters outside plain English turn into strings of odd characters.
Not in Markdown. The format has no way of joining cells across rows or columns, so no generator can offer it honestly. If you need merged cells, the HTML Table Generator gives you markup you can add colspan and rowspan to.
No. The table is built in your browser, and pasting, opening a file, copying and downloading all happen on your own computer. Nothing you put in the editor is sent to us or stored.