Privacy Redactor support
Last updated: 30 September 2026
Get started
Open Redact Prompts in ChatGPT. Download the models once, paste text into the private panel, review the detected highlights, and choose which details to keep. Click Redact, review the full output, then Send to chat or Copy.
Device requirements
Privacy Redactor currently requires WebGPU support and enough memory for its local models. Setup downloads approximately 1.1 GB of model files. Detection is optimized for English. It does not automatically intercept ChatGPT’s prompt box.
Troubleshooting
If setup fails, check your internet connection and WebGPU support, close memory-intensive apps, and try loading the models again. Clear the panel before sharing a screenshot. Do not send confidential text in a support request.
Contact
Use the contact form on the Procurato homepage. Include your operating system, browser or ChatGPT app version, and the error message. We do not need the original text to investigate a technical issue.