Free CSV viewers that open large files, compared
Most CSV viewers are fine until the file gets big. Then they either refuse to open it, load for ten minutes, or open it and quietly show you part of it. Here is what actually works, and at what size.
What “large” means here
- Under 100 MB. Almost anything works, including Excel.
- 100 MB to 1 GB. Excel struggles, text editors cope, most viewers slow down.
- 1 GB and up. Only tools that stream the file survive.
- 10 GB and up. Only tools that never load it at all.
The dividing line is whether the tool loads the whole file into memory. If it does, its ceiling is your RAM, roughly a third of it once parsing overhead is counted.
The free options
Text editors. Notepad++ on Windows (up to about 2 GB), Sublime Text, VS Code, EmEditor’s free tier. They open big files fast and show you the raw text. They do not give you columns, sorting, filtering or totals. Good for checking what is in a file, not for working with it.
LibreOffice Calc. Free, cross-platform, and it has the best import dialog of any spreadsheet: a type control per column before anything loads. It has the same 1,048,576 row limit as Excel and the same memory profile, so it is not a large-file answer.
Tad. Free, cross-platform, built on DuckDB. Opens large CSVs quickly and does pivots well. It is a viewer: you cannot edit cells.
csvkit and xsv, or qsv. Command line, free. xsv in particular handles multi-gigabyte files quickly for slicing, counting and sorting. You need to be comfortable in a terminal.
DuckDB. Free, and the fastest way to answer a question about a huge CSV if you write SQL:
duckdb -c "SELECT count(*) FROM 'bigfile.csv'"No import step, no row limit. The interface is a SQL prompt.
Online CSV viewers. Free and convenient. Do not use them for anything containing personal data. Uploading a file with names, emails, salaries or medical records to a third-party website is a data transfer, and in most organisations it breaks a policy that exists precisely to prevent it. There is no version of “I just needed to open it” that survives the conversation afterwards.
Quick comparison
| Opens 10 GB | Columns and sort | Edit cells | Free | |
|---|---|---|---|---|
| Notepad++ / VS Code | partly | no | text only | yes |
| LibreOffice Calc | no | yes | yes | yes |
| Tad | yes | yes | no | yes |
| xsv / qsv | yes | sort only | no | yes |
| DuckDB | yes | via SQL | via SQL | yes |
| Excel | no | yes | yes | no |
The gap in that table is the row that opens 10 GB, shows columns, sorts, and lets you edit, without requiring SQL. That is the gap BigXL was built for.
Which to pick
- Just need to see what is in it: a text editor.
- Need a total or a count, and you write SQL: DuckDB.
- Need to browse, sort, filter and fix values, without SQL: a streaming spreadsheet.
- File contains personal data: anything that runs on your own machine, and nothing that uploads.
What we built
BigXL is a free desktop CSV viewer and editor with no row limit, for macOS, Windows and Linux. It opens a 10 GB file in seconds because it never loads the whole thing: rows appear as you scroll, and sorting and filtering run through an embedded query engine over every row, not just the ones on screen.
Opening any size file, sorting, filtering, searching, editing cells, charts, pivots and CSV export are free, permanently, with no account and no sign-up. Nothing is ever uploaded, because the application makes no network calls at all.