Free Online Text File Viewer
Drop a .txt data file below and it opens as a table. Fixed-width column boundaries are found by reading the file, delimiters are detected by counting them, and what it decided is reported above the grid. Sort, filter, measure any column, and export as CSV. Nothing is uploaded.
The format that never went away
Fixed-width text is supposed to be extinct. It is not. Core banking systems, payroll bureaux, insurance clearing houses, government filings and anything that touches a mainframe still exchange data as text files where a field is not separated by a character but by a position: columns 1 to 12 are the transaction id, 13 to 24 the date, 25 to 46 the customer name, padded with spaces to fill.
These files are perfectly readable by the system that produced them and nearly unreadable by anyone else. A spreadsheet's import wizard will ask you to click the column boundaries by hand on a preview of the first few lines, which is fine for eight columns and awful for forty, and gets it wrong the moment one line is longer than the sample. The official answer is a copybook or a layout specification, which the person who sent you the file rarely includes.
There is a shortcut, and this page takes it. In a genuine fixed-width file, certain character positions are blank on every single line, because that is where the padding between fields lives. Find those vertical runs of whitespace and you have found the column boundaries, without a copybook and without clicking anything. It is what a person does by eye when they squint at the file; it is just faster in code and it looks at every line rather than the first twenty.
The same page handles the easier case too. If the file uses a delimiter, that is detected first by counting candidates per line and preferring whichever gives a consistent field count. Either way, the summary line above the grid tells you what it concluded, so you are never left guessing whether the table you are looking at is the table the file describes.
What is on the page
Fixed-width detection
Column boundaries found from vertical whitespace across the whole file, not from a preview of the first few lines. No copybook, no clicking on a ruler.
Delimiter detection
Comma, tab, semicolon, pipe, and runs of two or more spaces. Chosen by consistency of field count rather than raw frequency, so punctuation inside values does not win.
It tells you what it decided
The summary above the grid says fixed-width or which delimiter, and how many boundaries were found. A wrong guess is visible immediately instead of hiding in the data.
Padding trimmed, values untouched
The spaces that fill a fixed-width field are formatting and are removed. The value inside is not: an account number of 0044821 stays as written and is badged text rather than number.
Sort, filter, freeze
Click a header to sort, filter as you type across all columns or one, and pin the first columns so the record key stays visible on a wide extract.
Column statistics
Blanks, distinct count, min, max, mean, median, sum and most common values, one click from a header. Useful for checking a batch total before you pass the file on.
Export as CSV
Download CSV writes the current view with correct quoting, which is the whole job for most people: turn the extract into something a spreadsheet will open properly.
Big files stay smooth
Only the rows on screen become HTML, so a very long extract scrolls like a short one, and anything sizeable is read with a progress bar rather than a frozen tab.
Nothing leaves the tab
No upload endpoint, no account, no retention. Fixed-width extracts are usually payroll, banking or healthcare, and those are precisely the files that should not go to a web converter.
A worked example
The Try with sample data link loads a fixed-width ledger. The first lines look like this:
TXN_ID TXN_DATE CUSTOMER SKU DESCRIPTION AMOUNT STATUS
TXN-100241 2025-01-06 Acme Robotics 0042 Widget assembly 1,240.00 PAID
TXN-100242 2025-01-06 Northwind Traders 0117 Sprocket set 318.75 PAID
TXN-100243 2025-01-07 Contoso Ltd 0009 Gasket, 40mm 92.40 OPEN
Note what makes this hard for a naive reader. There is no delimiter. There are single spaces inside values (Acme Robotics), so splitting on whitespace would produce a different number of fields on every line. There is a comma inside Gasket, 40mm and another inside 1,240.00, so splitting on commas would be worse still. The only thing that is reliable is the position.
Dropped on this page, the summary reads fixed-width columns, boundaries detected at 7 positions and the grid has exactly the seven columns the header line names, with Acme Robotics intact in one cell and Gasket, 40mm intact in another.
The badges are worth a look. SKU reads text / id, not int, because the values have leading zeros; 0042 is a part code, and turning it into 42 would be a data loss you would not notice until someone could not find the part. AMOUNT reads num / currency despite the thousands separators, because the numeric reader understands them.
Click the sigma on AMOUNT for the batch total. Type STATUS:OPEN in the filter to see just the unpaid rows and read the count. Then Download CSV, and the file you get is properly quoted, so Gasket, 40mm arrives in a spreadsheet as one cell rather than two.
When detection is hard
- A field that is full on every line. If no line ever leaves a gap at a boundary, there is no gap to find and two fields will read as one. This happens with tightly packed numeric extracts. A copybook is the only real answer there.
- Ragged line lengths. Trailing fields that are empty on most lines produce a boundary the detection may place differently from the specification. Compare the column count against the header line; if they disagree, that is the tell.
- Report furniture. Page headers, footers, dashed rules and totals lines are not records, and they push the detection around. Files with a lot of that are better handled by the text to CSV tool, which lets you set the mode by hand.
- Tabs pretending to be alignment. A file aligned with tab characters is delimited, not fixed-width, and will be read that way. It usually looks wrong in a text editor and right here.
- No structure at all. A log file becomes one column of lines. That is the correct answer, and filtering and counting still work on it.
Frequently Asked Questions
How do I open a .txt data file as a table?
Drop it on the box at the top of this page, or paste the text into the Paste tab. The layout is worked out from the content: a delimiter if one is used consistently, otherwise fixed-width column boundaries found by looking for the character positions that are blank on every line. The result is a sortable, filterable table you can export.
How does it detect fixed-width columns?
By looking for vertical runs of whitespace. In a fixed-width file every field starts at the same character position on every line, which means certain columns of the file are blank all the way down. Those gaps are the boundaries. It is the same method a human uses when squinting at a mainframe report, and it works without a copybook or a layout specification.
What delimiters does it recognise?
Comma, tab, semicolon and pipe, plus runs of two or more spaces used as a separator. The choice is made by counting candidates per line and preferring the one that gives a consistent field count, so punctuation that appears inside values does not win. What it settled on is reported in the summary line above the grid.
Are my leading zeros and padding preserved?
Leading zeros yes, padding no. Fixed-width files pad fields with spaces to fill the column, and that padding is not data, so it is trimmed. The value itself is untouched: an account number of 0044821 or a SKU of 0042 comes through exactly as written, and a column of such values is badged text rather than number so nothing tries to renumber it.
What if the file has no columns at all?
You get a single column of lines, and the summary says so. That is the right answer for a log file or a list, and it is still useful: you can filter as you type, sort, count distinct values with the column statistics, and export. If the file has structure the detection missed, the delimiter is worth checking by eye first.
How big a text file can this open?
250 MB. Anything over 2 MB is read on a background thread with a progress bar so the tab keeps responding, and the grid only renders the rows on screen, so a very long file scrolls at the same speed as a short one. Above 250 MB, the full editor streams up to 1 GB.
Can I turn it into a CSV?
Yes, and that is the most common reason to open one. Download CSV writes the current view with proper quoting, so a fixed-width mainframe extract becomes a file a spreadsheet will open correctly. Download .xlsx does the same as a workbook, and for a fragment you can select a range and use Copy as.
Is the file uploaded?
No. There is no upload endpoint on this page. The text is read from your disk in your own tab and nothing is transmitted or stored. Fixed-width extracts are usually banking, payroll or healthcare data, which is exactly the category that should not be pasted into an online converter.
Related tools and guides
Open the extract nobody sent a layout for
Boundaries found from the file itself, values kept as written, and a properly quoted CSV on the way out.
Back to the viewer