Fast
GPU-accelerated transcription turns an hour of audio into text in a few minutes.
Unlimited audio & video transcription
Turn audio and video into accurate text in minutes.
Upload interviews, meetings, podcasts, lectures or videos and get a clean transcript with timestamps and speaker labels in 98+ languages.
- 3 free transcripts every day
- No credit card required
- Export to DOCX, PDF, TXT, SRT & VTT
98+languages
10 hper file
50files at once
Drag & drop audio or video
MP3, MP4, M4A, MOV, WAV, AAC, OGG, OPUS, WMA, MKV, WEBM and more
Browse files
Speaker recognition & more settings
By using ScribeFiles you agree to the terms and privacy policy.
Everything you need to go from recording to text
Built on Whisper, the most accurate open speech recognition model available.
Accurate
Three modes from Cheetah (fastest) to Whale (most accurate), so you choose the trade-off.
Speaker recognition
Every paragraph is labelled with who's speaking. Rename speakers with one click.
98+ languages, 130+ translations
Auto-detects the spoken language. Translate transcripts and subtitles into 130+ languages with the timings kept.
AI summaries & content
One click turns a recording into a summary, action items, chapters, a blog post or social posts. Or ask it anything.
Every export format
DOCX, PDF, TXT, CSV and subtitles (SRT, VTT), with or without timestamps and speakers.
Private by default
Only you can access your files and transcripts. Delete them whenever you want.
Edit while you listen
Click any sentence to jump to that moment, fix words inline and rename speakers. Changes save automatically.
How it works
- 1
Upload
Drop in audio or video files, up to 10 hours each.
- 2
Transcribe
Pick a language and mode. We handle the rest while you get on with your day.
- 3
Edit & export
Play back, fix any word, name the speakers, then download in the format you need.
Simple pricing
Start free. Upgrade when you need more.
Free
$0
Try it out, no credit card.
- 3 transcripts every day
- Files up to 30 minutes
- Upload 1 file at a time
- Cheetah & Dolphin modes
- Speaker recognition
- 1 AI summary or post a day
- Audio files kept 7 days
- Lower queue priority
Unlimited
$10 / month
$120 billed yearly
- Unlimited transcription
- Files up to 10 hours / 5 GB
- Upload 50 files at a time
- All modes, including 🐋 Whale
- Translate to 130+ languages & transcribe to English
- AI summaries, blog & social posts (20/day)
- Files kept until you delete them
- Bulk exports
- Highest queue priority
Frequently asked questions
What is ScribeFiles?
ScribeFiles is an AI transcription service that converts audio and video files to text in 98+ languages, with speaker labels, timestamps and exports for documents and subtitles.
How much does it cost?
The free plan includes 3 transcripts per day for files up to 30 minutes. Unlimited costs $10/month billed yearly, or $20/month billed monthly.
Is Unlimited really unlimited?
Yes. There's no cap on how many hours you transcribe. The only rule is that an account is for one person.
Which file formats are supported?
Nearly all common audio and video formats: MP3, M4A, MP4, MOV, AAC, WAV, OGG, OPUS, MPEG, WMA, WMV, AVI, FLAC, AIFF, MKV, WEBM, 3GP and more.
Can I upload large files?
On Unlimited, each file can be up to 10 hours long and 5 GB, and you can upload 50 files at a time.
Is my data private?
Your files and transcripts are only visible to you, and you can delete them, or your whole account, at any time. Payments are handled by Stripe; we never see your card number.
Can it translate or summarise my recordings?
Yes. Translate any transcript and its subtitles into 130+ languages, or use the AI tools to create summaries, action items, chapters, blog posts and social posts, or ask your own question about the recording.
How accurate is it?
On clear recordings, Whisper's accuracy rivals professional human transcribers. Background noise, crosstalk and heavy accents lower accuracy. Try Whale mode and "Restore audio" for difficult files.
Stop typing. Start transcribing.
3 free transcripts every day. No credit card required.
Transcribe your first file