Skip to main content

Human-grade transcripts, without the human price.

Every uncertain word is flagged for you — review the marks, not the whole transcript.

27 languages · from $0.50/min · 1 hr of audio ≈ 10 min

300 free minutes on a work or edu email · 60 on a personal one

interview_0729_participantK.m4a · 42:17In review · 2
04:12

Researcher: Where did you first come across the product?

04:19

Participant K: Adayafor you to confirm and I were on a field visit to Guanajuato last year — the distributors there call that stock "primera agua". Fast accents trip most tools.

2 to confirm · 1 verified against the web"Safe to export" unlocks at zero doubts

The problem

Three old problems with interview transcription.

Not accurate enough

Names, jargon, accents — other tools keep getting them wrong.

You can't see where it's wrong

One opaque pass — so you re-listen to everything.

Human help is slow and expensive

Human services charge per minute and take days.

Our answer

Review-first.

We never hand you one take-it-or-leave-it transcript. Every uncertain spot is pulled into a review queue for you to confirm — and only at zero doubts do we say "safe to export."

01Upload02Transcribe03Review04Export

Review only the marked spots — never re-listen to the whole tape.

Doubt04:19

"primera agua"

…the distributors there call that stock "primera agua"…

Engines heard: primera agua ×2 · primavera ×1

1 / 2

Why it's accurate

Several "stenographers" listen at once — then reconcile their notes.

For each language we pick the engines that hear it best, run them in parallel, cross-check every line, and verify key names and places against the open web. Where they disagree is exactly what gets flagged for you.

Stenographer Aprimera agua
Stenographer Bprimavera
Stenographer Cprimera agua

Fused transcript

…the distributors there call that stock "primera agua"…

A 2:1 split is never silently accepted — it's flagged for you.

Your project has its own vocabulary.

Names, product codenames, in-house jargon — put them in a glossary up front and the transcript comes back spelled your way, instead of needing a pass of find-and-replace.

Claiming accuracy ought to cost us something. If a transcription turns out to have a serious quality defect of our making, that transcription is refunded in full — how to raise it and how we review it is written into §4 of the Terms.

After review

A confirmed transcript can go two steps further.

The same pipeline, extended — one click from the detail page once doubts hit zero; charged only on completion.

Narrative

Q&A rewritten as a first-person narrative — reads like the participant telling it.

Redacted copy

Sensitive details replaced per your list — keep-listed words never touched — safe to share out.

Priced per minute, only for the steps you pick: narrative +$0.05/min · redaction +$0.05/min — checks built into each step, charged only on completion.

Pricing

In every language, about half the local human rate.

English

$0.50 /min

Other languages

$1.00 /min

Euro reference €0.45 / €0.90 (excl. tax)

Prorated to the secondRate locks at uploadFailed jobs never billed

See full pricing →

Site-wide to date

100,000+

hours of interview audio processed

2,000+

project glossaries built

27

languages, each individually tuned

100+

specialist fields covered

Every transcript is cross-checked by four engines running in parallel.The same audio is transcribed four times, and the four results are held against each other.
Flagship case · this case

Fortune Global 500 · telecoms equipment manufacturerGlobal customer satisfaction study

3,172 in-depth interviews40+ countries and regions18 languages

From raw audio to structured findings you can analyse — one pipeline, end to end.

See how this one was done →

A letter from the founders

An interview is the most expensive raw material in research.

— The founding team, Transcribe.

Try it on your own interviews.

300 free minutes on a work or edu email (≈ 5 hours of interviews); 60 on a personal one.

Start free →
Prorated to the secondFailed jobs never billedSerious quality defects refundedAudio auto-deleted after 7 days · transcripts kept 30 daysBalance terms: §4 of the Terms

FAQ

Common questions.