LOCAL FILES. CLEAR RESULTS.

CSV data cleaner

Diagnose a dirty CSV and clean it with auditable steps: a full local profile of every column, a suggested fix plan you approve item by item, and deterministic local execution with a change log.

About this tool

Every imported CSV is dirty in its own way: empty cells, exact duplicates, amounts stored as "$1,200.00", three date formats in one column. Downstream tools fail on all of it. This cleaner starts with a full local profile of your file — per-column types, empty rates, unique counts and duplicate rows — computed over every row, not a sample.

The model sees only that profile, never your data, and proposes a cleaning plan. You tick the steps you approve — trim, dedupe, drop empty rows, fill blanks, normalize amounts and dates, lower-case email columns — and deterministic local code executes them. Every step lands in a change log with affected counts, and your original file is never modified.

How to use CSV data cleaner

  1. Upload a CSV (up to 10 MB) or paste CSV text — it is profiled entirely in your browser.
  2. Add your provider API key in the key panel.
  3. Review the per-column diagnosis and the suggested cleaning plan, tick the steps you want.
  4. Apply the ticked steps locally and download the cleaned CSV with its change log.

Keep the original files and inspect the exported result before publishing or sharing it.

Frequently asked questions

What does the AI actually see?

Only the column profile — names, types, empty rates, value counts and up to 3 sample values per column. Your full data is never sent; cleaning itself runs locally with deterministic code.

What can it fix?

Trimming, exact duplicate rows, fully empty rows, filling empty cells, normalizing amounts like "$1,200.00" to plain numbers, converting dates to ISO, and lower-casing columns like emails.

Can I undo a cleaning step?

The change log records exactly what each step changed (how many cells or rows, in which column), and your original file is never modified — download a fresh copy and re-run with different steps any time.

How is this different from the AI data analyst?

The analyst answers questions about a CSV. This tool is a cleaning pipeline: profile, plan, apply, audit.

Related tools