Excel to Markdown converter

Drop a Excel spreadsheet below and get clean Markdown, with the token count and what it costs to send. It is converted in your browser — the file never leaves your device.

97% found their conversion helpful · 31 ratings

Drop a file here, or click to choose one

Converted in your browser — the file never leaves your device

or paste text

Input format

Output style

Rates checked 2026-08-31. See the cost on all 35 models → · Use this as context for a prompt →

What a XLSX actually stores

This is the conversion with the widest quality gap against pasting raw, and the reason is worth understanding. Fifty rows of comma-separated values give a model no reliable way to know which column a value belongs to once rows get long or contain commas. The same fifty rows as a pipe table make that unambiguous on every single line.

Spreadsheets are where the numbers live, and "explain this data" is one of the most common things anyone asks a model. Getting the table structure across intact is most of what decides whether the answer is right.

What survives the conversion

Cell values, resolved properly through the shared-string table so text arrives as text rather than as the numeric indexes Excel stores internally. Every worksheet is converted, each as its own section. The first row becomes the table header.

What does not

Formulas are not evaluated — you get the stored result, which is what you want anyway. Formatting, colours, merged cells and charts are all lost, because none of them survive as text. Very large sheets are capped at 500 rows per sheet, which is well past the point where summarising beats pasting.

Saying so plainly is deliberate. A converter that quietly drops half a document is worse than one that tells you which half, because you can work around a limit you know about.

Questions

Are formulas calculated?

No — the last calculated value stored in the file is used. That is almost always what you want, since a model cannot execute a formula either.

What about multiple sheets?

All of them convert, each becoming its own section with a heading, so the model can tell where one dataset ends and the next begins.

Why not just paste CSV?

You can, and it works. A pipe table simply reads better to a model on wide or long data, because the column a value belongs to is restated on every row rather than inferred from position.

Converting a different format

The same engine handles eighteen formats in total, all of them in your browser.

PDF → MD DOCX → MD PPTX → MD HTML → MD CSV → MD SRT → MD JSON → MD EPUB → MD IPYNB → MD EML → MD XML → MD LaTeX → MD Text → MD RTF → MD ODT → MD VTT → MD YAML → MD All 18 formats →

Token rates checked 2026-08-31. See the cost on all 35 models →

More free tools

All free, all instant, none of them need an account.

Prompt Scorer Paste a prompt, get a score and the specific gaps. Token Counter Token estimate plus what it costs on every major model. Framework Finder Describe your task, get the frameworks that fit it. Image Prompt Builder Subject, style, lighting and negatives, formatted per platform. Prompt A/B Compare See exactly which version is stronger, and where. System Prompt Builder Role, boundaries, tools and refusal rules in one block. AI Writing Check Flags the phrasing that makes text read as machine-written. Which AI Model? Match a task and a budget to the model that fits.