Fabraix Red Team Playground
Free · Open source
AI security researchers
PII GUI
FreeOpen SourceSelf-Hosted
Paste a contract into an AI tool and every name, phone number, and case ID in that file goes straight to the model — and its training pipeline. PII GUI sits between your documents and the AI, stripping identifiers on your machine before anything leaves.
Similar tools
Fabraix Red Team PlaygroundFreeFabraix Red Team Playground is an open-source, MIT-licensed testing environment structured…
EclusiaFreemiumEclusia runs locally on your machine and replaces names, IBANs, phone numbers, API keys, and…
Patronus ScannerFreemiumPatronus AI positions itself as evaluation and monitoring infrastructure for production LLM…
ArgotFreeArgot runs statistical checks against your repository's own commit history to flag code that…
Get This Tool
License: License: unverified
Local-run terms: Users can download, install, and run the desktop app locally on Mac, Windows, or Linux with no vendor involvement after initial setup.
Screenshots 1
Pricing
- Model
- Free
Summary
The app runs detection locally using on-device models, so nothing is uploaded at any point — no sign-up, no server round-trip, no cloud dependency. You review every flagged item in context before committing to a redaction, which means you catch the false positives before they become permanent holes in a legal document. Custom regex lets you add patterns the model won't know: internal case IDs, account number formats, bespoke identifiers. The export produces a PDF with sensitive text actually gone, not layered over. Where it breaks: single-file, single-session workflow with no batch processing described in the docs, so teams processing hundreds of support logs daily will hit a throughput ceiling fast.
Bottom line: The right call for anyone who needs to clean a contract or medical record before it touches an AI tool — not the right call for a team that needs to process a folder of five hundred files without touching each one.
Find your fit
Before you choose PII GUI
More options in Guardrails & Safety. Compare pricing, platforms, and features side by side.
Eclusia
Freemium
Legal professionals handling confidential cases
Pros
- On-device detection with local models, so documents never leave the machine — which means you can process medical records or legal files that contractually cannot touch a third-party server.
- Inline review before any redaction is committed, so you catch the false positives that a blind auto-redact would permanently remove from a contract.
- Custom regex support for account numbers, case IDs, and proprietary identifiers, so the model's blind spots don't become your compliance gaps.
- Export produces PDFs with text genuinely removed rather than covered, so a downstream recipient cannot recover the original content by manipulating the file.
- No account, no sign-up, and no trial expiry, so the tool is available when you need it without an approval cycle or a billing conversation.
Cons
- No batch processing is described anywhere in the docs or page content — the workflow is one document opened and reviewed at a time. A team processing hundreds of support logs daily will be clicking through files manually, and at that volume they move to a scripted pipeline built on an NLP library like spaCy or Presidio instead.
- No API surface is available, so redaction cannot be inserted into an automated document ingestion workflow. Any team that needs redaction to happen programmatically — before files hit a storage bucket, for example — cannot use this tool as-is and will need a self-hosted server-side solution.
- The local model downloads on first use, which means the first run on an air-gapped machine or a machine with restricted outbound access requires planning. The docs describe it as a one-time download, but teams in strict network-controlled environments need to account for that step.
About
- Platforms
- Mac, Windows, Linux
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- June 21, 2026
Best For
Who it's for
- Users needing offline PII redaction
- Reviewing and controlling redactions line-by-line
- Exporting permanently cleaned PDFs
- Adding custom regex for specific identifiers
What it does well
- Redact PII from emails before sharing with AI tools
- Clean contracts and legal documents for sharing
- Process medical records or HR files offline
- Prepare support logs without exposing personal data
Compare PII GUI
Frequently Asked Questions
- Is PII GUI free?
- PII GUI offers a free option. Hosting, model usage, or optional services may have separate costs; check the pricing details and official terms.
- Is PII GUI open source?
- Yes. PII GUI is open source.
- Can I self-host PII GUI?
- Yes. PII GUI supports self-hosting on your own infrastructure.
- What platforms does PII GUI support?
- PII GUI is available on: Mac, Windows, Linux.
People also compared
Curated lists that include this category
Most AI workflows assume the document going in is safe to share. PII GUI addresses the case where it isn’t: you open a PDF, text file, or Markdown document, the on-device model runs detection, and every flagged item — names, emails, phone numbers, IDs — is highlighted in place for your review. You toggle detections on or off, add your own regex patterns for identifiers the model won’t recognize, and export a redacted copy. The local model downloads once; after that, the tool runs with no internet connection required.
The review step is what separates this from a black-box redaction script. Detection results are surfaced inline — you see exactly what the model flagged and in what context before anything is removed. That matters for legal and medical documents where a wrongly redacted clause or a missed identifier both carry real consequences. You stay in the loop; the tool doesn’t decide for you.
The vendor describes the app as built for contracts, medical records, HR files, and support logs — documents that need to be shared without exposing the people named in them. The free desktop build is available for Mac, Windows, and Linux, and the source is open. The documented workflow is one document at a time, in one window, on one machine. Teams that need to push batches of files through a pipeline, integrate redaction into an automated process, or expose the functionality via API will find no supported path for that here — and will likely end up running a self-hosted open-source NLP pipeline instead.
The app requires no account and no login. The detection model downloads on first use and operates offline from that point forward. Supported input formats are text, Markdown, and PDF; output is a redacted PDF or text file with sensitive content removed rather than visually obscured.
Contribute
- Work on PII GUI? Claim this listing.
- Embed a badge for your README or site.
Sign in to rate this tool or suggest an edit.