Every word you can trust. Every doubt flagged.
Upload an interview, meeting or podcast and get a speaker-labeled transcript in minutes. Words the engine wasn't sure about are highlighted with likely alternatives, so checking an hour of audio takes minutes, not an afternoon.
Built for people who live in recordings
Most transcripts look confident. Ours tell you where to look.
Every engine makes mistakes. The difference is whether you have to reread everything to find them.
Uncertain words, flagged
Low-confidence words are highlighted in amber with the engine's next-best guesses. Tab between them and press 1, 2 or 3 to fix.
Who said what
Speakers are separated automatically. Rename them once and every turn updates; reassign a line with Alt+number.
Audio stays in sync
Click any word to hear it. The playhead follows along as you read, and rewinds a second when you resume so you never lose your place.
Summary and chapters
Each transcript comes with a short summary, chapter markers, action items and keywords, so you can skim before you dig in.
Email a file, get a transcript
Every project has its own address. Forward a recording from your phone and the transcript lands in your inbox when it's done.
An API from day one
Create jobs, upload audio and fetch transcripts over a plain REST API with signed webhooks. Same price as the app.
Proofread an hour of audio without touching the mouse.
The editor was designed around the way professional transcriptionists work: hands on the keys, audio under your control, and a jump straight to the next thing that needs attention.
- Jump between flagged words and pick an alternative in one keystroke
- Play, pause, skip and change speed without leaving the text
- Split and merge turns, insert [inaudible] and [crosstalk] tags
- Find and replace across the whole transcript, with undo
Send it anywhere, in the format they actually want.
Export to Word, plain text, subtitles or JSON, with or without timestamps and speaker names. Or send a read-only link that plays the audio alongside the text and expires when you say so.
- Every save is versioned, so you can compare and restore
- Invite editors and viewers to a shared workspace
- Set how long audio is kept, per project
Three steps. Most files finish before your coffee does.
Upload or email a file
Drag in audio or video (MP3, M4A, WAV, MP4, MOV and more), paste a link, or email it to your project address. You see the exact price before anything is charged.
We transcribe and flag
Speakers are separated, punctuation is added, and every word the engine wasn't sure about is marked with alternatives. You get an email when it's ready.
Review, then share
Tab through the flagged words, fix what's wrong, and export or share. You're only charged for the minutes of audio we transcribed.
One price. Pay only for what you transcribe.
Buy credit in packs, and it never expires. No seats, no tiers, no surprise renewals.
That's $1.80 for an hour-long interview, billed to the second, deducted from prepaid credit.
- $3 free credit when you sign up (100 minutes)
- Bonus credit on $50 and $100 packs
- Editor, summaries, exports, sharing and API included
- Monthly spending cap and low-balance alerts
| Typical alternatives | Same audio |
|---|---|
| Human transcription (~$1.50/min) | $450 |
| Subscription apps (seat + overage) | $30+/seat |
| Scribberty | $30.00 |
Good to know
How accurate is it?
On clear recordings, typically above 95% of words are right before you touch anything. What matters more is that the words most likely to be wrong are the ones we flag, so your review time goes where it counts.
Can I use it for medical or health information?
Not yet. Scribberty is not HIPAA compliant today and we do not sign Business Associate Agreements. Please do not upload recordings that contain protected health information (PHI), such as patient names together with diagnoses, treatment or billing details. A HIPAA-ready option is on our roadmap.
Which languages and file types do you support?
English is fully supported, with automatic language detection for more than 30 other languages. We accept most audio and video formats, including MP3, M4A, WAV, FLAC, OGG, MP4, MOV and WEBM, up to 2 GB per file.
How long does it take?
Most files finish in a fraction of their running time; an hour of audio usually takes a few minutes. You'll get an email when it's done.
What happens to my audio?
Your files are stored encrypted and are only used to produce your transcript. They are never used to train models. You choose how long audio is kept per project, and you can delete any file at any time.
Do unused credits expire?
No. Prepaid credit stays on your account until you use it.
Your next transcript is 100 free minutes away.
No card needed to start. Upload a file and see the flagged words for yourself.