Keep this tab open — you'll be redirected to your transcript.
MP3WAVM4AMP4MOVMKV+ more
Why this transcriber is actually free
Almost every transcription tool gates your transcript behind an account, a watermark, or an email form. This one does not: anonymous visitors get 3 files a day, up to 30 minutes and 100 MB each, with full TXT, SRT, and VTT exports. We run our own GPU servers, so short files cost us almost nothing to process — the free tier is the product, and paid plans for long files (up to 10 hours) pay for it.
A free account raises the limits to 5 files a day at up to 1 hour each and saves your transcript library. Pro adds priority processing, speaker labels, and DOCX/PDF exports.
Built for real recordings
The engine is a Whisper-class large speech model — the same family of models used by professional transcription services. It handles meetings, lectures, podcasts, interviews, voice memos, and phone recordings in 90+ languages with automatic language detection, and it punctuates and cases text properly so the transcript reads like writing, not like subtitles glued together.
After processing you land in a transcript editor: click any line to fix it, search the text, toggle timestamps, then export in one click.
Frequently asked questions
Is it really free? What's the catch?
The free tier is exactly what it says: 3 files a day per visitor, up to 30 minutes and 100 MB per file, with no account, no watermark, and no email gate. The catch is only the per-file length cap — longer files (lectures, podcasts, full meetings) need a free account (1-hour files) or Pro / a credit pack (10-hour files). That's the whole business model.
Do I need an account?
No. Drop a file and you get the transcript and all three core exports (TXT, SRT, VTT) without signing up. An account is optional: it raises the daily and per-file limits and keeps your transcripts in a library instead of auto-deleting them.
What happens to my file after I upload it?
Anonymous uploads and their transcripts are automatically deleted 24 hours after upload — treat the transcript link as temporary and export what you need. We never use your audio to train models, and files are not shared with any third party; processing happens on our own GPU servers.
How accurate is the transcription?
On clear speech with a decent microphone, a Whisper-class large model is near-human — typically a low single-digit word error rate in English and other well-supported languages. Accuracy drops with heavy background noise, crosstalk, strong accents, or very quiet recordings, which is why every transcript opens in an editor for quick fixes.
How long does it take?
Usually a fraction of the audio length: a 10-minute recording typically transcribes in about a minute once processing starts. At busy times your file may briefly queue — the progress page shows the live status, and Pro jobs skip to the priority queue.
Which file types can I upload?
Audio: MP3, WAV, M4A, FLAC, OGG, AAC, WMA, Opus, AIFF, AMR. Video: MP4, MOV, MKV, WebM, AVI, WMV, MPEG, 3GP. For video we read only the audio track, so resolution and file size matter far less than duration.