Launched this week
Hide emails, SSNs, card numbers and API keys in a screenshot before you share. One-click auto-redact plus drag-to-blur — 100% in your browser, nothing uploaded, no account. And a full screenshot studio too: frames, backgrounds, transparent PNG and batch export. Free, or $24.99 once for 4K.
How does the auto-redact actually detect sensitive info accurately, especially for things like custom API key formats? Curious how much fiddling it needs before it just works on the first try.
Hey@sedatyeew great question, this is the part I obsess over most.
Auto-redact runs OCR on your screenshot locally (nothing leaves the browser), then pattern-matches the text for the things with a recognisable shape: emails, phone numbers, IBANs, card numbers, and API keys/tokens. For keys it leans on high-entropy detection plus the common prefixes (sk-, ghp_, AKIA, Bearer…), so the standard formats from the big providers get caught on the first pass with zero setup.
Where it can't read your mind is a truly custom key that looks like ordinary text — no prefix, low entropy.
I don't want to pretend that's magic: that's what the one-click drag-to-blur is for, and the blur is baked into the exported PNG, not a layer anyone can peel back. So worst case it's "auto catches the obvious ones, you drag over the one weird one."
Honestly, custom key formats are exactly what I most want to improve — if you share the shape of one that slipped through, I'll add it. That kind of feedback is gold right now. Thanks for digging in 🙏
"Nothing gets uploaded" so how does the AI background removal / OCR actually run? Is it really local?
@ryan_storteboom
Yep, genuinely local. The AI background removal and the OCR behind auto-redact both run client-side in your browser — the image and the recognised text never get sent to a server, because there is no server in the loop for editing. No account, no upload, no signup. That's the whole reason I built it this way: framing a screenshot is throwaway-frequent, and shipping your screen contents to someone else's cloud for that felt wrong. You can literally go offline after the page loads and it still works.
Drag-to-blur for credit card numbers is genuinely useful, saved me from manually pixelating in Preview. Loved that it stays fully local—felt way less sketchy than uploading to some random site.
@furkanbeki61828 Thank you this genuinely made my day. "Saved me from pixelating in Preview" is exactly the moment I built it for. And you nailed why it's local: for something as throwaway as a screenshot, shipping your screen contents off to some server always felt backwards to me, so there just isn't one in the loop — the image never leaves your device.
Since you've clearly got the redaction use-case dialed in: does auto-redact catch the sensitive stuff for you on the first pass, or are you mostly reaching for the manual drag-to-blur? That's the bit I most want to get right, so any rough edge you hit is gold. Really appreciate you trying it 🙏