Bounded memory
Index and stream where practical instead of materializing multiple full copies of the table.
CSVProof is designed to stay useful when a CSV outgrows a comfortable spreadsheet workflow, without making an unverified gigabyte claim the product story.
Private early access · macOS & Windows
Opening a large file is only the first checkpoint. A credible large-file workflow must also search, filter, compare, edit, save, recover and verify the output without hiding skipped records.
CSVProof will publish limits only after repeatable tests on named hardware, versions and fixtures. Current product previews are not performance benchmarks.
The methodology reports time, memory and correctness together. A fast open that silently drops data is not a pass.
Planned corpus tiers: 100 MB, 1 GB, 5 GB and 10 GB on documented Mac and Windows hardware. Results will separate direct observations from vendor claims.
Index and stream where practical instead of materializing multiple full copies of the table.
Show a usable grid and parser state before every background task has finished.
Journal edits, write a new output atomically, and verify counts and hashes at completion.
Modern CSV, SmoothCSV and Caxton already make credible large-file claims or show strong performance. CSVProof’s reason to exist is the controlled path from parser contract to exceptions, repeatable recipe and evidence, not “largest file wins.”
CSVProof is currently in private early access for macOS and Windows.
Get early access