📊 CSV Viewer
Paste or upload a CSV file and view it as a sortable, filterable table. Auto-detects delimiter and handles quoted fields.
Last updated: November 8, 2026 · By the BoltQuickTools team
Looking inside a CSV without opening a spreadsheet
You have an export or a log and want to see its columns, sort it and find a few rows, without importing it into Excel and without Excel reformatting your IDs and dates. This page shows it as a table. The file is read with the browser's FileReader and parsed by the page's own script, with no upload step in the code.
Viewing a file
- Paste into the text box, click Upload CSV for a
.csv,.tsvor.txtfile, or click Try a sample. - Leave Delimiter on Auto-detect or choose Comma, Semicolon, Tab or Pipe. Untick First row is header if your data has no header, and columns become
col1,col2and so on. - Click a column heading to sort. Click it again to reverse.
- Type in Filter rows for a case-insensitive search across every column.
- Download with Export JSON or Export Clean CSV.
With the sample data
The sample has five rows, two with commas inside quotes, such as "Chen, Wei". The status line reads:
5 rows shown (5 total) · 4 columns · delimiter: ,
Clicking salary twice gives Eli 105000, Chen, Wei 91000, Bob 82000, Alice 75000, Diana 68000. Typing tx in the filter leaves only Diana, in Austin, TX, and the status changes to 1 rows shown (5 total).
Export JSON writes data.json. Every value is a string, even the numbers:
[
{
"name": "Alice",
"age": "30",
"city": "New York",
"salary": "75000"
},
...
]
Parsing and sorting surprises
- Only the first 5,000 matching rows are drawn. A note tells you when more exist. The row counts and exports still include everything, and the whole text is parsed again on each keystroke in the filter, so very large files will feel slow.
- ISO dates do not sort properly. The sort reads the leading number from each cell, so
2026-03-15and2026-01-02are both just 2026 and keep their original order. Years sort, months and days do not. Values like12abcsort as 12. - Extra cells disappear. The header row sets the column count. In a row with more cells, the extras are neither shown nor exported. Short rows are padded with blanks.
- Repeated header names collapse in JSON. Two columns called
abecome one key holding the last value. - Blank lines become empty rows and are counted.
- An unclosed quote runs on. A stray
"absorbs following lines into one cell with no warning. If the row count looks low, look for one. - Auto-detect needs at least two lines. A single line is always split on commas.
- Files are read as UTF-8. A file saved in an older encoding such as Windows-1252 will show broken accented characters. Re-save it as UTF-8.
Exports, clearing and conversions
Do exports respect my filter and sort?
No. Both buttons write all parsed rows in their original order, whatever the table is showing. To export a subset, cut it down in your source first.
What does Clean CSV actually change?
It switches the delimiter to a comma, adds or removes quotes as needed, and pads or trims rows to the header's width. It does not trim whitespace inside cells, so " x " keeps its spaces.
Why did Export still download data after I clicked Clear?
Clear empties the box and table, but the last parsed data stays in memory until you load something new. Your CSV is not sent anywhere, but do not assume the export will be empty.
Can I go between CSV and other formats?
For more control over JSON, use the JSON to CSV converter and check the result in the JSON formatter. To paste a small table into a README, try the Markdown table generator.