TrueTranscribe
True verbatim transcription with pauses, mm-hmms and crosstalk
About TrueTranscribe
TrueTranscribe is a speech-to-text service built for true verbatim transcripts. Instead of returning only the words, it records the whole conversation: pauses with their length, short acknowledgements such as mm-hmm said over another speaker, moments when people talk at once, and laughter and sighs. Every transcript has speaker roles and a timestamp for every word. That level of detail matters for research interviews, therapy and coaching sessions, legal and HR conversations, and any work where how something was said carries as much weight as what was said.
The service runs speech recognition on its own servers, and the site states that recordings are never sent to third-party services. Recognition is available in 14 languages, including English, Hindi, Spanish, Arabic, French, Portuguese, Russian, Indonesian, German, Turkish, Italian, Polish, Ukrainian and Dutch. Pricing is pay as you go by the minute of audio, with three turnaround speeds and no subscription. New accounts get their first transcription free, up to 120 minutes, plus $5 of credit.
Sign up and upload a recording. Choose how fast you need it: within 24 hours for archives, within 2 hours for same-day analysis, or instant processing at the front of the queue. When the transcript is ready, open it in the editor. Rather than rereading everything, you press Next doubtful spot to jump to the places where recognition may have gone wrong, and the snippet plays straight away. Words the recognizer is unsure of are underlined, and lines where the voice does not clearly match the speaker are boxed. Fix a word in the text, reassign a line to the other speaker, split or merge lines, or insert a pause or an mm-hmm. Lines you have listened to or corrected are marked as reviewed. When you are done, export the transcript to PDF or DOCX in one click.
- •True verbatim output - Pauses with durations, overlapping speech, back-channel sounds, laughter and sighs, with speaker roles and per-word timestamps.
- •Review-focused editor - Uncertainty highlighting, a Next doubtful spot button, listening from any word (slowed down if needed) and one-gesture fixes, also usable on a phone.
- •Learning from corrections - A personal dictionary fixes a word automatically after you correct it the same way twice, and your terms are fed to the recognizer in advance.
- •Personal speaker settings - The service learns which short overlaps you mark as simultaneous speech and, with consent, recognizes your own voice in new recordings. Other speakers' voices are not stored.
- •Own-server processing - Recognition runs on TrueTranscribe's servers rather than third-party APIs, and training data is deleted along with your profile.
- •Per-minute pricing - Three turnaround speeds from $0.04 to $0.10 per minute, with PDF and DOCX export.
The site names transcription agencies, coaches, psychologists and psychotherapists, researchers, UX researchers and marketers, journalists and podcasters, lawyers and mediators, clinics, HR and recruiters, teachers and speech therapists, sales and customer service teams, and managers recording meetings. Qualitative researchers and UX teams get transcripts that keep hesitations and interruptions for analysis. Therapists and coaches can review sessions with speaker roles intact. Transcription agencies can use the editor to check machine output quickly before delivering it. Because there is no subscription, it suits people with irregular volumes who would rather pay only for the minutes they transcribe.
Pricing
Usage based, no flat monthly plan
Pay per audio minute, no subscription: $0.04/min (within 24 hours), $0.06/min (within 2 hours), $0.10/min (instant). First transcription free up to 120 minutes, plus $5 on your account at sign-up.
From the vendor pricing page, 2026-11-09
