Roadmap

Here now

Every one of these runs in your browser today. Nothing about them is a preview.

• Merge two CSV files by a common column — Combine two spreadsheets by matching a shared column, like VLOOKUP. Handles a million rows, and your file never leaves your computer.
• Remove duplicate rows from a CSV — Find and remove repeated rows — exactly, or by close match with every group shown before anything goes. Your file never leaves your computer.
• Compare two CSV files and see what changed — See exactly which rows were added, removed, or changed between two exports, down to the individual cell. Nothing is uploaded.
• Check a CSV file for problems — Drop a file and see what is wrong with it: blanks, repeated rows, garbled characters, mixed date formats, numbers stored as text. Nothing is uploaded.
• Edit a CSV file — without anything else changing — Click a cell and fix it. Leading zeros, dates, and long numbers stay exactly as written, and your file never leaves your device.
• Find and replace in a CSV — Replace text across the whole file or one column, exact or by pattern, with a preview of every change before you download.
• Remove blank rows from a CSV — Strip empty and whitespace-only rows in one pass, with a count of what went and a preview before anything is downloaded.
• Fix a semicolon CSV — the European export, made standard — Turn a semicolon, tab or pipe separated European export into standard comma CSV — decimal commas converted so 1.234,56 stays the number it is. No upload.
• Sort a CSV file by a column — Sort a CSV by any column, as text or as numbers, blanks last. Every value stays exactly as written and the file never leaves your device.
• Filter the rows of a CSV by a value — Keep only the rows where a column equals, contains, or is empty — and see the count before you download. Nothing is uploaded.
• Take a random sample of rows from a CSV — Pick a random 100 rows, or 5 percent, from a CSV of any size — for a test file, a spot check, or a survey — with nothing uploaded.
• Keep only the first or last rows of a CSV — Cut a CSV down to its first 1,000 rows, or its last 50, with the heading kept — a quick test file from a big export, in your browser.
• Count the rows in a CSV file — How many rows does the file really have? Counted properly — quoted line breaks and all — for a file of any size, without opening it in a spreadsheet.
• Shuffle the rows of a CSV — Put the rows of a CSV in random order, heading kept — for a fair draw, a blind review, or a training file — in your browser, nothing uploaded.
• Delete columns from a CSV — Tick the columns to drop and download the rest, byte for byte unchanged. Useful before sharing a file that holds more than the recipient needs.
• Transpose a CSV — turn rows into columns — Turn the rows of a CSV into columns and the columns into rows, with the first column becoming the new heading row. In your browser, nothing uploaded.
• Split one CSV column into several — Split a column on a separator — a space, comma, dash, or any text — into as many new columns as it needs, with the original kept or dropped.
• Combine two CSV columns into one — Put two or more columns together into one — first and last name, street and city — with the separator you choose, and see it before you download.
• Add a row number column to a CSV — Add a numbered column to a CSV — starting from 1 or any number, as the first or last column — so every row has an id. In your browser, nothing uploaded.
• Convert CSV to Excel — without your numbers changing — Turn a CSV into an .xlsx workbook in your browser. Leading zeros, long numbers and dates stay exactly as written, and the file never leaves your device.
• Convert Excel to CSV — one clean file per sheet — Turn an .xlsx workbook into clean CSVs: every sheet listed, merged cells unmerged, title rows detected, formulas as their values. No upload.
• Convert CSV to JSON — records, columns, or rows — Turn a CSV into JSON in your browser: an array of records, one array per column, or row arrays. NDJSON too. Nothing is uploaded, nothing is retyped.
• Convert JSON to CSV — nested fields flattened, nothing guessed — Turn JSON or NDJSON into a CSV: records, column and row shapes all read, nested objects flattened to dotted columns, numbers kept exactly as written.
• Convert CSV to SQL INSERT statements — Turn a CSV into INSERT statements with correct quoting for your database, an optional CREATE TABLE, and batched rows. Runs entirely in your browser.
• Convert CSV to a Markdown table — Paste-ready Markdown tables from any CSV: columns aligned so the source reads as plain text, pipes and line breaks escaped. Runs in your browser.
• Convert TSV to CSV — with the quoting done right — Turn a tab-separated file into a comma-separated one, quoting the values that need it. Runs entirely in your browser; nothing is uploaded or retyped.
• Convert Excel to JSON — a sheet as an array of records — Turn an .xlsx sheet into JSON records in your browser: merged cells handled, title rows skipped, formulas as values, nothing retyped.
• Convert CSV to Parquet — columnar, compressed, local — Turn a CSV into an Apache Parquet file inside your browser, written by a real analytical engine. No upload, no server, no signup.
• Convert Parquet to CSV — open the file your tools sent you — Read an Apache Parquet file in your browser and save it as CSV. Powered by a real analytical engine; nothing is uploaded anywhere.
• Extract a table from a web page or pasted HTML — Paste the HTML of a page, or the table you copied from it, and get the rows as CSV: merged cells split out, headers kept, nothing sent anywhere.
• Run SQL on a CSV file — Drop files, they become tables, write SQL against them. A real analytical database running in the page, with nothing sent anywhere.

In the plan

Specified and on the way, but not built yet — so they are named here and nowhere linked. A page whose buttons did nothing would be worse than no page at all.

• Combine many CSV files into one soon — Stack many files into a single one, even when their columns do not quite line up. Mismatches are reported rather than silently dropped.

Later — the one thing we would ever charge for

A paid tier is coming, and it covers convenience, never the tools: saving a clean-up as a recipe and re-running it on next month's export, and workspaces that live on your device. If you do the same job every month it will be worth paying for; if you do not, the free tools already do everything they do today.

It is not built yet, and the pricing page says so plainly until it is. When it ships, the free tools stay exactly as free as they are now.

What we will never do

These are not roadmap items that keep slipping. They are lines the product will not cross.

• Upload your data to any server, for any reason — including "just for processing." The work happens on your device or it does not happen.
• Ads, trackers, or third-party analytics with a payload. There is nothing watching what you do here.
• Require an account for any free tool, or put a row limit or a watermark on one.
• Sell, share, or derive anything from your data. There is nothing to sell — that absence is the product.

You do not have to take our word for any of it. Check it yourself — cut your network connection and the tools keep working, because there was never a server in the loop.

Shape it

The plan above is a hypothesis. If the tool you need is missing, or the one you use is missing something, tell us — what you ask for is what moves up.