# Audio redaction: redact spoken PII, names, and IDs

Automatically silence spoken names, phone numbers, emails, addresses, and IDs in audio and 4K video.

Powered by [Bleep censor](https://audo.ai/bleep-censor.md) — automatic AI PII detection and custom terms silenced in place.

## How it works

1. Upload a call, interview, or video (WAV, MP3, M4A, AAC, MP4, MOV, MKV, and more) up to 10 GB.
2. Leave Auto-detect PII on All PII (or pick No names to keep people's names), and choose Silence or Tone.
3. Review starred matches, untick any to keep via Redact only these, and download.

## Where teams run automatic audio redaction

Silence spoken identifiers in place without changing duration or re-encoding video.

### Support & QA calls: Automatic PII redaction for customer calls

Catch caller names, street addresses, SSNs, card numbers, and spoken digits like "five five five, oh one nine nine" or spelled-out emails.

### Interviews & podcasts: Keep names while redacting contact details

Pick No names to leave first and last names in a podcast or [Interview transcription](https://audo.ai/interview-transcription.md) clip while muting phone numbers, emails, addresses, and IDs.

### Companies & codenames: Combine AI detection with custom wildcards

Add up to 200 company names, codenames, or acme* prefix wildcards under Also redact alongside automatic PII detection.

### 4K video & legal records: Tone markers and untouched video

Pick Tone for an audible 1 kHz public-records marker while copying 1080p and 4K MP4, MOV, MKV video bit-for-bit.

## AI PII detection with quiet-valley audio edges

Powered by Audo and its [Bleep censor](https://audo.ai/bleep-censor.md) engine, this audio redaction software makes it simple to redact audio and video automatically — finding spoken personal names, phone numbers, emails, street addresses, and IDs across 60 languages on audio (WAV, MP3, M4A, AAC, FLAC, OGG, Opus, WebM, AIFF) and video (MP4, MOV, MKV, WebM, AVI) up to 10 GB, with the first minute free per file, 3 times a day, and no sign-up.

Full-sentence context catches spoken digits, spelled-out "at" and "dot" emails, SSNs, and account numbers automatically, plus optional acme* wildcards. Audo places edges in the quiet valley up to 120 ms outside each word with 8 ms fades, caps 1 kHz Tone at −3 dBFS, copies 1080p or 4K video untouched, and stars every match for verification or [Auto captions](https://audo.ai/auto-captions.md).

Choosing the right audio redaction mode on Audo

| Redaction mode | What gets silenced or toned out | Rate & best use |
|---|---|---|
| All PII (default) | Spoken personal names, phone numbers, emails, street addresses, and IDs/numbers across 60 languages | $0.04/min ($0.03 min) — customer support calls, QA reviews, and public records |
| No names | Phone numbers, emails, street addresses, and IDs/numbers while keeping people's names intact | $0.04/min ($0.03 min) — keeps names in interviews, podcasts, and user research |
| Off + Also redact | Up to 200 custom companies, codenames, phrases, or acme* prefix wildcards | $0.03/min — custom companies, codenames, and acme* prefixes without AI PII scanning |
| Keep the background | Vocal stem separation so music and ambience keep playing under each redacted word | +$0.05/min add-on; re-rendering ticked hits via Redact only these is $0.01/min (+$0.05/min with background) |

Tool: Censor audio (`censor_audio`). The same tool's other pages: https://audo.ai/bleep-censor.md, https://audo.ai/clean-song-maker.md. Price: 3 credits a minute — $0.03/min ($1.80/hr), or $0.01/min ($0.60/hr) with transcript_file_id, +$0.01/min (+$0.60/hr, $0.03 paid-run min) with pii, +$0.05/min (+$3.00/hr) with keep_background, pay-as-you-go with no monthly subscription — of input for whole files, charged only when a job succeeds.

## Free API

```bash
curl -F file=@support-call.mp3 -F pii=all -F profanity=none -F sound=mute https://audo.ai/api/try/censor-audio -D response.headers -o response.body
```

Read response.headers and response.body before naming media. Handle 202, 200 JSON and errors with the complete response client at https://audo.ai/docs/api?utm_source=markdown&utm_medium=agent#responses.

Free for the first minute of each file, 3 times a day per person, with no key or account. Uploads up to 10 GB, 2 at once from one address. The response is the result itself; a job that runs longer than about 30 seconds returns 202 with a link to check on it. With an API key (`-H "Authorization: Bearer $AUDO_API_KEY"`), omitting free runs the full job with credits if the balance covers it, otherwise the free allowance if quota remains. Set free=true to prohibit spending; for paid runs use free=false with an approved positive max_credits. A free result cut short comes with finishUrl (and the Audo-Finish-Url header when the response is the file itself): give it to the person, who pays there (from $5, no sign-up), and the whole file runs; keep checking the status link, whose finishedBy names the whole-file job and its own status link. Details: https://audo.ai/docs/api.md

## Options

Form fields with the same names as the MCP tool's inputs:

- `profanity`: Which swear words the built-in list censors: strong (strong swear words and slurs), all (also mild ones such as damn, hell, and crap), or none (only words). Lists for en, es, fr, de, pt, it, nl. Default strong.
- `pii`: Personal details (PII) to find automatically in speech: all (names, phone numbers, emails, street addresses, and IDs/numbers), contact (phone, email, address, and IDs/numbers, keeping people's names), or none. Adds 1 credit a minute (at least 3 credits a paid run). Default none.
- `words`: More words or phrases to censor, in any language, up to 200, such as ["acme", "john smith"]. Matched as whole words, ignoring case and punctuation; end one with * to match every word that starts with it, such as "frick*".
- `allow`: Words the built-in list would censor that should stay, such as ["hell", "bloody"]. Default: none.
- `sound`: What a censored word sounds like: bleep (a 1 kHz tone at the voice's own level) or mute (silence). Default bleep.
- `keep_background`: true separates the voice first, so music and sound effects keep playing under each censored word and only the voice is bleeped or muted. Adds 5 credits a minute. Default false: everything is bleeped or muted for the word.
- `dry_run`: true lists the words it would censor, with their IDs and times, and makes no file. Default false.
- `censor_ids`: Censors only these, by ID from a dry run with the same transcript, such as ["w12", "w48"]. Needs transcript_file_id. Default: everything found.
- `language`: The spoken language, as a code such as en or es. It picks the swear word list: en, es, fr, de, pt, it, nl. Default: detected.
- `output_format`: The result's format: same (as the input; a video stays a video), wav, mp3, m4a, or flac. Default same.

## Languages

Transcription recognizes 60 languages, so automatic AI PII detection (`pii: "all"` or `"contact"`) and custom words and phrases in `words` (plus `allow` lists and pre-starred transcript words) work in all 60 languages. Built-in swear-word and slur lists (`profanity: "strong"` or `"all"`) cover 7 languages: `en` English, `es` Spanish, `fr` French, `de` German, `pt` Portuguese, `it` Italian, and `nl` Dutch.

## From an assistant (MCP)

Add https://audo.ai/mcp and sign in (setup for each assistant: https://audo.ai/setup.md). The tool is `censor_audio`:

> Bleeps or mutes swear words, slurs, spoken personal details (PII), and words or phrases you list, in place: length, timing, and a video's picture stay. pii "all" or "contact" automatically detects spoken personal names, phone numbers, emails, street addresses, and IDs/numbers ("contact" keeps names). Profanity lists for en, es, fr, de, pt, it, nl; your words work in any language. Edges go in the quiet, so no syllable slips through. keep_background keeps music and effects playing under each word. dry_run lists matches with IDs. Not for cutting words out (cut_audio). It costs 3 credits a minute, or 1 with transcript_file_id, plus 1 a minute with pii (at least 3 credits a paid run) and 5 with keep_background, charged only if the job succeeds. Returns the file, each word's times, and a transcript marking them, or job_id for get_job.

Example: "Use Audo to automatically redact all personal names, phone numbers, and IDs in support-call.mp3."

## Questions

**Is there free audio redaction software with automatic PII detection?** Yes, Audo redacts the first minute of each file for free, 3 times a day, with no sign-up and AI PII detection included. Whole files with automatic PII redaction cost $0.04 a minute ($2.40 an hour) (at least $0.03 per paid run), custom-word-only redaction is $0.03 a minute ($1.80 an hour), re-rendering ticked matches from the saved transcript is $0.01 a minute ($0.60 an hour) (plus $0.05 a minute ($3.00 an hour) when Keep the background is on), and Keep the background adds $0.05 a minute ($3.00 an hour), paid once from $5 with no subscription — [see pricing](https://audo.ai/pricing.md).

**What personal details does Audo detect automatically, and can I keep names?** Audo finds spoken personal names, phone numbers (even spoken digits like "five five five, oh one nine nine"), emails (even spoken "at" and "dot"), street addresses, and ID, SSN, account, or card numbers across 60 languages. Switch to No names to keep people's names in interviews while muting private numbers and addresses, or add custom acme* wildcards under Also redact.

**How do I verify redacted matches or restore a word?** On Audo, every match appears starred in its sentence with a timestamp and category tag (Name, Phone, Email, Address, or ID / number). Signed in, untick any match and press Redact only these to re-render from the saved transcript for $0.01 a minute ($0.60 an hour) (plus $0.05 a minute ($3.00 an hour) when Keep the background is on).

**Does video redaction keep 4K quality and lip sync?** Yes — Audo replaces audio samples in place with 8 ms fades while copying 1080p or 4K video bit-for-bit without blurring faces or drifting lip sync, and [Auto captions](https://audo.ai/auto-captions.md) masks redacted words as stars.

**Can developers or AI agents automate PII redaction?** Yes, pass pii="all" or pii="contact" to censor_audio via Audo's [REST API](https://audo.ai/docs/api.md), or [Add Audo to your assistant](https://audo.ai/setup.md) for Claude, ChatGPT, Claude Code, and Cursor.

## More

- Other cuts: [Remove filler words](https://audo.ai/remove-filler-words.md), [Bleep censor](https://audo.ai/bleep-censor.md), [AI clip finder](https://audo.ai/ai-clip-finder.md), [Cut audio](https://audo.ai/audio-cutter.md), [Remove silence](https://audo.ai/silence-remover.md), [Trim video](https://audo.ai/trim-video.md), [Clean song maker](https://audo.ai/clean-song-maker.md)
- Related guides: [Audio and video tools for AI agents](https://audo.ai/guides/audio-and-video-tools-for-ai-agents.md)
- [All tools](https://audo.ai/all-tools.md)
- [Prices and free limits](https://audo.ai/pricing.md)
- [Set up your assistant](https://audo.ai/setup.md)
- Every tool: https://audo.ai/llms.txt
- OpenAPI: https://audo.ai/api/openapi.json
- This page for people: https://audo.ai/audio-redaction?utm_source=markdown&utm_medium=agent
