Hey Product Hunt 👋 I built csvdiff.app because I kept doing the same dumb dance: export two CSVs, open both in Excel, squint at them side by side, and inevitably miss the one row that changed. Existing diff tools treat CSVs like text files — so a reordered column or an extra quote mark lights up the whole file as "changed" when nothing meaningful did. I wanted something that actually understands tabular data: match rows by key, show me what was added, removed, or modified, and ignore the noise. It grew a bit beyond diffing — you can filter, sort, and poke around your data right in the browser. Everything runs client-side. Your files never leave your machine, no upload, no account, no telemetry. That was non-negotiable since I use this with work data myself. It's free. Would love to hear what's missing — especially if you work with exports from tools that produce cursed CSVs. I've seen some things, but I'm sure you've seen worse.

Uploaded a messy csv with about 8,000 rows and the duplicate detection caught inconsistencies I had missed manually. The export to a cleaned file saved me a good chunk of time on a small client project.