Redact PII
Hosted LLM strips personal data — names, emails, phones, addresses, SSNs, card-shaped numbers — replacing each with [REDACTED]. Catches free-form mentions ("met Joe at Acme HQ in March") that pattern-based redaction misses. Uses 2 credits per run.
About Redact PII
Redact PII strips personal data out of text, replacing names, emails, phone numbers, addresses, SSNs, and card-shaped numbers with [REDACTED]. Because it's powered by a hosted LLM, it also catches free-form mentions like "met Joe at Acme HQ in March" that pattern-based redactors slide right past. Use it before sharing logs, transcripts, or notes that might contain sensitive identifiers; each run costs 2 credits.
- Category
- privacy
- Input
- Accepts: text/plain.
- Output
- Outputs: text/plain.
- Cost
- Credit-metered
- Memory
- low
Common uses
- Scrub customer names and emails from a support transcript before pasting it into a ticket
- Remove addresses and phone numbers from notes before sharing them in a public channel
- Strip SSNs and card-shaped numbers from a document headed for a less-trusted system
- Sanitize a chat log of contextual mentions like a person met at a named company
- Clean meeting notes of personal identifiers before archiving or distributing them
- Prepare a sample dataset of free-text fields for sharing without exposing real people
Frequently asked questions
What kinds of data does it redact?
Names, emails, phone numbers, addresses, SSNs, and card-shaped numbers, each replaced with [REDACTED]. It also catches free-form mentions of people and places that regex misses.
Why is it better than pattern-based redaction?
Pattern matching only catches structured formats; the hosted LLM also recognizes natural-language mentions like a name dropped mid-sentence.
Is my text uploaded for redaction?
Yes. The redaction is performed by a hosted model, so your text is sent to it. For fully on-device redaction, use the free in-browser Text Redact tool instead.
What does it cost and what does it return?
2 credits per run. It takes plain text and returns the same text with personal data replaced by [REDACTED].
Is redaction guaranteed to catch everything?
It catches far more than pattern-based tools, but no automated redactor is perfect; review sensitive output before sharing it widely.
Keywords
- redact
- pii
- privacy
- mask
- sanitize
- pro
- llm