Verbatim transcripts for journalism.
Quote people accurately. Label every speaker in multi-speaker recordings.
1,158 minutes of journalism recordings already transcribed here · first transcript free, any length · ~2 min per hour of audio · 99 languages · speaker labels on every file · private, never used for training
- 1,158 minof journalism recordings transcribed by users here, across 30 files
- ~2 minper hour of audio, measured on production
- 99languages, speaker labels on every one
At a glance
- Field
- Journalism
- Transcribed here
- 1,158 minutes across 30 journalism recordings, out of 146,086 minutes site-wide since April 2026
- Turnaround
- About 2 minutes per hour of audio; most files under a minute
- Input
- Any audio or video file, YouTube, podcast or RSS link; up to 10 hours or 5 GB
- Output
- Speaker labels, word timestamps, summary, AI chat; TXT, SRT, VTT, DOCX, JSON
- Languages
- 99, with automatic detection
- Privacy
- Private servers, no third-party AI, never used for training, delete any time
- Price
- First transcript free at any length; second $0.99; then credit packs from $8 for 1,000 minutes, never expire
Built for Journalism
Four things a reporter actually needs from a transcription tool.
Verbatim fidelity
Ums, overlaps, and "uh"s stay in. You can clean afterwards, but the source-of-truth output is faithful. Word-level timestamps let you hear any moment before you quote it.
Multi-speaker labels
Speaker 1, Speaker 2, Speaker 3 labels on every turn. Rename once, the whole transcript updates. Works on Zoom exports, phone recordings, and single-mic room captures.
Off-the-record tagging
Highlight any range and tag it off-the-record. Excluded from exports by default. Your editor sees clean copy; your notes keep everything.
99 languages
Arabic, Hindi, Mandarin, Spanish, French, Portuguese, Ukrainian, Russian — 99 via Whisper. Transcribe in source language, optionally output an English translation alongside.
How it works
Drop it, read it, export it.
Drop or paste
A file, a YouTube link, a podcast link or an RSS feed. Up to 10 hours or 5 GB.
Read in a minute
Speaker labels, timestamps, a summary and AI chat over the transcript. Click any line to hear it.
Export
TXT, SRT, VTT, DOCX or JSON. Every file stays in your library, deletable any time.
Questions
FAQs for journalism workflows.
Is Whipscribe verbatim or cleaned up?
Verbatim by default. Ums, uhs, false starts, and overlapping speech are preserved. You clean or redact afterwards; the source transcript stays faithful.
Can you handle multi-speaker interviews?
Yes. Speaker labels mark Speaker 1, Speaker 2, etc. on every turn. Rename once; the whole transcript updates.
How do I mark something off-the-record?
Tag a range as off-the-record in the editor — it stays visible to you but excluded from exports by default. Delete files + audio anytime.
Record once. Quote accurately.
Your first transcript is free at any length; leave an email to open it. Packs from $8 for 1,000 minutes after that; credits never expire.