Stop leaking secrets to AI.
Redact catches credentials and PII before they reach ChatGPT, Claude, Gemini and more. Runs in your browser.
Built for the chats you actually use
The paste box is a credential leak waiting to happen.
Chat transcripts sit on the provider's servers. You can't take it back.
Three steps. Every paste.
Redact intercepts pastes on supported chat sites. Detection runs in your browser.
01
Detect
A local neural network scans each paste for credentials and PII. Inference runs in about 150 milliseconds.
02
Decide
Set Block or Warn per data type from the popup. Defaults are tuned for engineers.
03
Paste
Send a redacted version, send the original, or cancel before it reaches the chat.
Privacy
Paste stays in your browser.
Detection runs locally. The extension makes no network calls. Every claim below is verifiable in the source.
-
On-device inference
The model runs in a Web Worker inside the browser. Paste content is processed locally.
-
No telemetry
Redact contains no analytics or error-reporting SDK. The source contains no outbound network calls.
-
Model bundled in the extension
The model file ships inside the extension package and loads via chrome-extension:// at startup.
-
Auditable
The source is available on GitHub under the PolyForm Noncommercial license.
Works on every major LLM chat.
Active only on these chat sites. Want another? Open an issue on GitHub.
ChatGPT
Claude
Microsoft Copilot
Perplexity
Grok
DeepSeek
Mistral Le Chat
Poe
Install once. Catch every paste.
Free in the Chrome Web Store with source available on GitHub.