Home / Guides / Open a huge CSV

Large-file workflow

How to open an 8 GB CSV without importing it first.

Use a viewer that indexes the file on disk, confirm its shape, then narrow the data before exporting anything downstream.

Fast path: Open the file directly in csvlite, wait for the row count, inspect delimiter and encoding, then filter or profile the columns you care about. The source remains read-only.

Why ordinary viewers struggle

An 8 GB CSV can require much more than 8 GB of memory once every cell becomes an object in an editable grid. Spreadsheet row limits can intervene before memory does. File-backed viewers instead index record boundaries and decode visible cells as needed.

A reliable checklist

  1. Keep the file on a fast local SSD when possible.
  2. Open it directly—avoid converting it merely to view it.
  3. Check delimiter, encoding, header detection, row and column counts.
  4. Jump to the end and inspect a few edge rows.
  5. Apply a selective filter before profiling or exporting a subset.
  6. Record the original file hash if the result matters operationally.

What about Mac?

The current csvlite release supports Windows and Linux, not macOS. On a Mac, DuckDB is a strong local option if SQL is comfortable, and Modern CSV offers a graphical application. We would rather state that limitation plainly than turn “view huge CSV Mac” into a misleading download page.

Observed csvlite timing

On one 13th Gen Intel i7 laptop with 32 GB RAM and NVMe storage, an 8.0 GB, 50-million-row, 15-column file opened in 4.3–8.5 seconds. This was an observed development run, not a controlled cross-product benchmark; hardware, file shape and cache state matter.

Open the file you actually have.

The app works directly with CSV, TSV, pipe/semicolon-delimited files and JSONL.

Get csvlite from $29