1. First-pass draft
Automated speech recognition prepares a first draft for our editors to work from.
How it works
From recording to transcript, step by step
Upload, see the price, pay, and receive a transcript that people have edited and checked. No account, no surprises after payment.
1. Upload and see your price
You pay for the work your recording actually needs, and you see why before you pay.
Measured
Each recording is measured to the second. You pay per audio minute, rounded up once per order, never per file.
Graded A to D
Every recording is graded on how hard it is to transcribe: how clearly the words come through, background noise, recording level, distortion and overlapping speakers.
Shown before you pay
Your quote lists each recording's grade and the reason. Think a grade is wrong? Email us before paying and a person checks it.
Accurate to the audio
What's written is what was said. If a transcript falls short, we re-edit it free, and every grade gets free corrections for 14 days. How the guarantee works
- Grade A · ClearClose microphone, little noise, one voice at a time.Standard rate
- Grade B · FairSome noise or distance, still easy to follow.Standard rate
- Grade C · DifficultNoisy room, distant speakers or frequent crosstalk.+30%
- Grade D · Very difficultHeavy noise, distortion or long unclear stretches.+70%
2. Pay online
Your quote holds for 7 days. Pay by card or M-Pesa on the upload page or from the link in your quote email. Prices are in US dollars, and work starts as soon as the payment is confirmed.
3. Three stages, two people
2. Professional edit
An editor listens to the whole recording, corrects words, names and terms, and labels every speaker.
3. Separate quality check
A second person who didn't edit the file checks it against the audio before it reaches you.
Read more about the automated steps on our How we use AI page.
4. Delivery and corrections
Transcripts arrive by email in 2–3 business days, or within 24 hours with rush delivery, as a branded PDF, DOCX or TXT; captions as SRT or VTT.
Transcripts are accurate to the audio: what's written is what was said. If one falls short, send us examples within 14 days. A reviewer who didn't work on your order checks it, and we re-edit it free within 48 hours. If that still isn't right, we look at a refund case by case. See the refund policy.
How your files are protected
- Encrypted connections for every upload and download, and encrypted storage at rest.
- Only the people working on your order can open your files; team tools need their own credentials.
- Audio and video deleted 90 days after upload, or sooner if you ask.
- Your recordings and transcripts are never used to train AI models, and our automated steps run with providers configured so they can't keep or train on them.
- Card and mobile-money details are entered on our payment provider's secure checkout. We never see card numbers or PINs.
- Confidentiality agreement (NDA) on request when you upload.
We hold no security certifications yet. Our Security page lists what is in place and what is still in progress.