Accurate transcripts, reviewed by a human.
TranscriptScribe pairs fast speech recognition with a human reviewer who listens, corrects, names every speaker and checks your terminology. You get back an accurate transcript you can share, file or publish.
How it works
1. Upload
Upload audio or video up to 6 hours long and choose a turnaround. Add names and terms we should get right.
2. AI Analysis
Speech recognition produces a first draft with timestamps and a label for each speaker.
3. Human review
A reviewer listens along, fixes errors, names the speakers and checks every term on your list.
4. Delivery
We email you when it is ready. Download Word, PDF, plain text, or SRT and VTT captions.
Built for work where accuracy matters
- Legal and compliance: interviews, hearings, depositions and recorded calls.
- Research: qualitative interviews and focus groups, with every speaker identified.
- Media: podcasts and video, with captions ready to upload.
- Business: board meetings, earnings calls and minutes you need to stand behind.
Simple per-minute pricing
| Turnaround | Price |
|---|---|
| 5 days | $1.25 per audio minute |
| 2 days | $1.75 per audio minute |
| 24 hours | $2.50 per audio minute |
Questions
How accurate is it?
Every transcript is read against the audio by a qualified person before it reaches you. Accuracy still depends on the recording: heavy crosstalk, strong background noise or very quiet speakers can leave words that cannot be made out.
Which files can I upload?
Any common audio or video format, such as MP3, M4A, WAV, MP4 or MOV, up to 2 GB and 6 hours.
What if something is wrong?
Reply to your delivery email within 7 days and we will correct it at no charge.
Is my recording kept private?
Files are stored privately and are only seen by the reviewer working on your order. Read our Privacy Policy for the details.
