A raw transcript or recording → the same words, cleaned: fillers and false starts removed, punctuation fixed, broken into paragraphs.
Files are encrypted
EXAMPLE
ALL-HANDS.M4A — RAW
arrow_forward
ALL-HANDS-CLEAN.TXT
so um the the launch is, like, ready to go
The launch is ready to go.
and uh we we moved the date to march third
all-hands-clean.txt · same words, cleaned text
What you get
check_circleFiller words, false starts, and stutters removed
check_circlePunctuation and capitalization fixed throughout
check_circleText broken into readable paragraphs
check_circleEvery sentence kept — nothing paraphrased, summarized, or reordered
check_circleWorks on a recording or on a transcript you already have
How it works
1Upload a recording or a transcriptDrop in one audio or video file up to 500 MB, or a transcript you already have as .txt or .docx.
2We transcribe if needed, then cleanRecordings are transcribed on our own servers first. Then a language model removes fillers and false starts, fixes punctuation and casing, and paragraphs the text — without rewording it.
3Download the clean transcriptA plain-text file and a formatted DOCX, both named after your upload, holding the same words as the source — just cleaner.
Who uses this tool
PODCASTING
Episode transcripts without the ums
Publish a transcript that reads like writing while staying word-for-word what was said on the show.
JOURNALISM
Interview quotes, ready to lift
Clean the fillers and fix the punctuation so quotes can go into the piece without retyping.
TEAMS & MEETINGS
Meeting transcripts people will read
Turn a raw machine transcript into paragraphs the team can actually skim — with nothing left out.
EDUCATION
Lecture transcripts as study text
A cleaned lecture reads like course notes but keeps every point the lecturer made, in order.
UX RESEARCH
Interview data you can code
Fillers gone, sentences intact — the participant's actual words survive for analysis.
CONTENT & VIDEO
Talks turned into readable text
Clean up a webinar or conference talk transcript before it becomes an article or captions.
SECURITY
Built for sensitive documents
Bank statements, medical files, case records. Security isn't a feature we added — it's the foundation.
lock
Encrypted in transit
Encrypted in transit over TLS 1.2+, and stored in access-controlled, encrypted object storage. Files are protected the moment they leave your browser.
admin_panel_settings
Your files stay yours
Workspaces are isolated per account. Role-based access shows teammates only what they need.
auto_delete
Deleted, not stored
Files are deleted after processing. Everything runs on our own hardware and is never sent to an outside AI service, so your data is never used to train models.
dnsProcessed on our own hardwareauto_deleteDeleted after processinglockEncrypted in transit
Questions
A recording — audio like MP3, M4A, WAV, AAC, FLAC, OGG, OPUS, WMA, or video like MP4, MOV, M4V, WEBM, MKV, AVI — or a transcript you already have as .txt or .docx. Recordings are transcribed first, then cleaned. One file per job, up to 500 MB.
No. This is a clean-verbatim edit: fillers, false starts, and stutters come out, and punctuation and casing are fixed, but the speaker's own words, order, and meaning stay. Nothing is paraphrased, summarized, reordered, or dropped — every sentence in the source comes back out.
No. If you upload a recording, the transcript has no speaker labels or timestamps. If the transcript you upload already contains them, they are kept — cleanup never drops content.
6 credits per minute of audio. Because the length is not known until the file is processed, the upfront hold is estimated from file size at a typical bitrate for the format (about 1 MB per minute for MP3 audio, more for WAV, FLAC, or video). The minimum is 6 credits — and since text files are tiny, .txt and .docx uploads usually land at that minimum.
Yes. Files run on our own servers, are never used to train any model, and are auto-deleted 7 days after the job finishes.