anonymizer
English · Русский
Offline

Offline PII redaction — nothing leaves your machine

anonymizer is a free, open-source tool that strips names, companies, IDs, addresses, emails and phone numbers from .docx, .pdf and .xlsx — entirely on your computer. No upload, no cloud, no telemetry. Built for lawyers and teams who can't send confidential documents to a hosted service.

Why offline matters

Cloud redaction tools upload your document to someone else's server — a non-starter under NDA, attorney-client privilege, or GDPR / 152-ФЗ. anonymizer processes everything locally, so the file never leaves your control. The no-network guarantee is enforced by a CI test, not just a promise.

What it does

Handles .docx, .pdf, and .xlsx — document structure and formatting preserved, metadata cleared. Detects English / EU PII alongside Russian identifiers: ИНН, СНИЛС, passport numbers, Cyrillic names and addresses. Every detected entity is replaced with a structured reversible token — hover a token in the in-browser preview to see the original value.

vs cloud tools

anonymizer is the only option that combines free & open-source (AGPL), fully offline, native multi-format support, and both Russian and English entity detection (plus an optional Spanish pack) in a single tool. See the full comparison.

Install

$ uv tool install docs-anonymizer

New to it? The ChatGPT workflow guide walks through a full run.

FAQ

Does it really run offline?
Yes. Detection runs locally with Natasha + spaCy + rules. A build-time test asserts the engine opens no network socket — a regression fails CI before release.
Does anything get uploaded to a server?
No. Documents are processed on your machine and never sent anywhere. There is no telemetry.
What does it redact?
Names, companies, addresses, emails, phones, and financial/government IDs (incl. Russian ИНН, СНИЛС, passport numbers) in .docx, .pdf and .xlsx — replaced with structured, reversible tokens. Spanish names, companies and addresses are detected too when the optional Spanish pack is installed.
Is it free?
Yes. AGPL-3.0 — no fees, no subscription, no seat limits. Install with uv tool install docs-anonymizer.

Get started

$ uv tool install docs-anonymizer

Full documentation — CLI reference, format coverage, detector configuration — at anonymizer.site/docs.