CSV Viewer
Open or paste CSV and read it as a real table: automatic delimiter detection, sortable columns, a filter across every cell, per-column type profiling and export to CSV or JSON.
Reviewed 17 November 2026 · Free · No account needed
How to use
- Paste CSV text or open a .csv/.tsv file — the delimiter is detected automatically, and you can override it.
- Toggle 'First row is header' if the file has no header row, or if the first row is real data.
- Click any column header to cycle ascending, descending and unsorted; use the filter box to search every cell at once.
- Column profile shows the inferred type, filled and unique counts, plus min/max/average/sum for numeric columns.
- Export the filtered view as CSV or copy it as a JSON array of objects.
Example
Input: name,role,started,salary Ava,Designer,2023-04-10,72000 Noah,Engineer,2022-09-01,88000
Output: A 2-row table with numeric profiling on the salary column (min 72,000 · max 88,000 · avg 80,000)
How it works
The CSV is parsed with Papa Parse, which handles quoted fields and embedded commas correctly. The result is shown as a table you can scroll and search.
Papa Parse reads the text with automatic delimiter detection (comma, semicolon or tab) and RFC 4180 quoting, so commas inside quoted fields stay intact. The first row becomes the table header and the row and column counts are reported alongside the detected delimiter.
"name,role / Ava,designer / Sam,dev" renders as a 2-column table with 2 data rows and comma reported as the detected delimiter.
Edge cases to know
- The preview shows up to 200 rows; the counts cover the whole file.
- The first row is always treated as the header — header-less CSVs will show their data as headings.
- Parse problems are reported per row; the first few are shown.
Limits & compatibility
- Up to 25 MB of CSV text is parsed; beyond that the input is clipped with a visible warning.
- Pagination keeps the table fast — the 'All' option renders at most 5,000 rows at once.
- Type inference is a heuristic: values like 007 or 1,234 are treated as text/number respectively based on the whole column.
- Quoted fields, escaped quotes and embedded newlines are handled by Papa Parse.
Privacy
Parsing, sorting and profiling run in your browser. Your data — including salaries, emails or customer records — never leaves the device.
Related tools
Questions about CSV Viewer
How does it decide what separates the columns?
The delimiter is detected automatically among comma, semicolon and tab, and the detected value is displayed above the table so you can confirm it. Quoted fields are parsed to RFC 4180 rules via Papa Parse, so commas inside quotes and escaped quotes behave as a spreadsheet would expect.
My file has tens of thousands of rows — will they all render?
The table shows the first 200 rows to stay responsive, but the row and column counts cover the whole file, so you still learn its true size. Very wide files scroll horizontally rather than being truncated. Paste is the intended input; parsing runs locally with nothing uploaded.
What if my data has no header row?
The first row is always treated as the header, so a header-less CSV will display its first data row as column names — something to check when the first column looks like values rather than labels. Parse errors are reported too: up to three messages show above the table when the input cannot be parsed cleanly.