CSV Data Profiler
Drop a CSV file to get a data quality report in seconds: what each column contains, how complete it is, its range of values, duplicates and anything unusual. Large files are streamed in chunks, so even a few hundred megabytes work.
Or paste CSV text
How to use this tool
- Drop a CSV file (up to 200 MB).
- Read the overview and the quality notes.
- Open a column to see its statistics and most common values, then download the profile as CSV.
How it works
Every row is read once. Each cell is classified as empty, number (1,234.5 or 1234,5), date (ISO or day/month/year), email, yes/no or text; a column gets a type when at least 95% of its filled cells share it, otherwise it is "mixed". Duplicate rows are exact matches of all cells. Mean and standard deviation use all numeric cells (sample standard deviation). Unique counts stop being exact above 50,000 distinct values per column.
Limitations
- Day/month and month/day dates cannot always be told apart; dates are read as day/month/year unless written as ISO (YYYY-MM-DD).
- Exact duplicate detection is limited to the first 2 million distinct rows.
Privacy and data
Files are read and processed in your browser. They are never uploaded to our server or anyone else, and nothing is kept after you close the page.