Encrypted on your device

Advanced AI transcription for audio and video.

GDPR-compliant, client-side encrypted and processed on secure EU infrastructure. Privacy without the fine print.

Audio becomes text. Fast.

Upload audio or video files and scryp creates a transcript automatically. Fast, accurate and available in more than 130 languages.

  • 1 hr audio in under 5 min
  • Language detection (134 languages)
  • Audio & video up to 10 GB per file
TumorBoard_Vienna.mp3

Tumor board AKH Vienna

47 min · 108 MB · MP3

Done
Ready
Encryption
Upload
Transcription
Speakers

18,432

Words

5

Speakers

99.7%

Accuracy

Who said what? Labelled automatically.

Different speakers are separated automatically and assigned to the right sections. Rename one passage, everything from that point on, or all matching speaker labels at once.

  • Automatic speaker diarization
  • Color-coded speaker labels
  • Rename speakers with one click
Speaker 1Speaker 2
00:00

Welcome. Today we're talking about the Vienna startup scene. Great to have you here.

00:05

My pleasure. There's an incredible amount going on in Vienna right now.

00:09

Absolutely. You founded your company in Seestadt. How did that come about?

00:14

I was at the technical university before and realized – research alone doesn't cut it. You have to get out there and actually build something.

Automatic demo – click to take over

Edit directly in your browser.

Correct passages directly in the transcript while listening to synchronized audio. Your changes are encrypted in your browser before they're saved.

  • Edit text inline with a click
  • Audio playback synced to the transcript
  • Corrections client-side encrypted
Saved & encrypted
Dr. Brosch
00:12

Recommendation: cholecystectomy for persistent symptoms.

Ms. Berger
00:18

I see. How long is the inpatient stay after the procedure?

Automatic demo – click to take over

Your transcript. Your rules.

Adapt every transcript to the way you read: font, speaker labels, timestamps or compact continuous text. One click updates the view.

  • Fonts: Sans, Serif or Mono
  • Toggle speakers & timestamps
  • Continuous text or segment view
Client_meeting_15-03.mp3
Mr. Wallner
0:00

So, Ms. Eder. The articles of association for your company are ready. Share capital of 10,000 euros, of which at least 5,000 to be paid in cash.

Ms. Eder
0:12

Very good. And what happens next?

Mr. Wallner
0:15

Next week we'll execute the notarial deed, then I'll file the commercial register application. After the registration number is assigned we'll apply for the trade license – expect six to eight weeks in total.

Ms. Eder
0:33

Perfect. I've already prepared the business plan and the bank confirmation of the capital contribution.

0:00Mr. WallnerSo, Ms. Eder. The articles of association for your company are ready. Share capital of 10,000 euros, of which at least 5,000 to be paid in cash. 0:12Ms. EderVery good. And what happens next? 0:15Mr. WallnerNext week we'll execute the notarial deed, then I'll file the commercial register application. After the registration number is assigned we'll apply for the trade license – expect six to eight weeks in total. 0:33Ms. EderPerfect. I've already prepared the business plan and the bank confirmation of the capital contribution.

Automatic demo – click to take over

Click. Listen. Land on the right moment.

Click any passage and the audio jumps to the exact moment it was spoken. The active segment is highlighted, so you can find every detail again in seconds.

  • Click text to jump to the matching audio
  • Live highlighting of the active segment
  • Play encrypted audio right in the browser
Podcast_Gynecology.mp3
0:00 / 0:26
Host0:00

How do you deal with the shift away from the trusted local specialist towards an outpatient model in gynecology?

Dr. Gruber0:09

Well, that's really extremely critical to watch. Although it does of course come out of a certain pressure.

Dr. Gruber0:14

Because every door handle gets gilded, figuratively speaking.

Dr. Gruber0:20

And ultimately it isn't the hospitals' job to provide the outpatient care that primary care is meant to handle.

Automatic demo – click to take over

Every format. No compromises.

Export a formatted PDF for your archive, an editable Word document, subtitles for video or clean plain text. Every transcript is ready to download with one click.

  • PDF, Word, TXT, SRT and WebVTT
  • Speakers, timestamps and formatting included
  • One click – formatted and ready to download

PDF .pdf

Formatted document with speakers and timestamps

Word .docx

Editable document with structured paragraphs

TXT .txt

Text file with timestamps and speakers

Plain text .txt

Clean continuous text – no metadata

SRT .srt

Subtitles for video players and YouTube

WebVTT .vtt

Web subtitles with speaker tags (W3C)

High accuracy. Built for real speech.

Our SX-3 model is optimized for high-quality transcription across conversations, interviews, meetings and dictation. The result: fewer errors and transcripts you can work with right away.

  • Own AI model optimized for everyday and business speech
  • Handles accents, proper names and natural speech reliably
  • Strong SX-3 default model for accurate everyday and business transcripts
Transcription accuracy – English
scryp
-
OpenAI
-
Google STT
-
Azure Speech
-
AWS Transcribe
-
70%100%
scryp SX-3
Other providers

Clearer audio. Better transcripts.

VoiceClear automatically reduces background noise, hiss and interference for more accurate transcripts, even when the original recording is difficult.

  • AI-based noise reduction
  • Optimized for speech and dictation
  • Automatic with every transcription
VoiceClear demo
0:00 / 0:00

Loading audio files...

Background noise, hiss and interference audible

Hours of recordings.
Reduced to what matters.

Extract the essentials from any transcript automatically: meeting minutes, action items, summaries or social posts. You choose the output.

  • Minutes, to-dos, summaries or posts with one click
  • Structured and ready to use
  • Exclusive to the Ultra plan
Transcript
AI analysis
Summary
00:00 – 03:20 Welcome + agenda...
03:20 – 12:45 Q4 quarterly figures...
12:45 – 18:00 2026 budget planning...
18:00 – 24:30 HR topics + open points...
24:30 – 26:10 Next steps + date...

Your data belongs to you alone.

Your audio files are encrypted in your browser before they reach our servers. Transcripts are stored encrypted as soon as they are created. Our servers permanently hold encrypted content only.

Client-side encryption

AES-256-GCM encryption directly in your browser

Your key, your data

Only you hold the key – not even we can read stored content

EU infrastructure

Processing on secure EU infrastructure

Compliance

GDPR-compliant – no sharing with third parties

Encryption flowEncrypted
Audio
AES-256
0x7f…

Your browser

AES-256-GCM encryption

Our servers

Store only encrypted data

Protected

Only you hold the key

Even in a data breach, your transcripts are useless without your key

Transcription is our business. Not your data.

Many transcription services process recordings unencrypted on third-party infrastructure. With scryp, your data never leaves your browser unencrypted – and our servers permanently store encrypted content only.

0recordings used for AI training

Our models are trained exclusively with paid speakers and open sources like Mozilla Common Voice – never with your data.

0data points shared with third parties

No tracking, no analytics sharing, no data trading. Your usage remains private.

0employees with plaintext access

Our servers store encrypted content only. Without your personal key, it is unreadable to anyone.

0copies of your original files

After processing, your original is irreversibly deleted. Only the encrypted playback version remains on the server.

Frequently asked questions

How does the encryption work?

Your audio files are encrypted in your browser with AES-256-GCM before they reach our servers. Our servers permanently store encrypted content only.

What happens to my audio files?

Your audio files are encrypted in your browser before they reach our servers. For AI transcription, the audio is processed only in a protected environment. Your data is stored permanently encrypted – without your key it is unreadable to anyone.

Which languages are supported?

Our AI model supports 134 languages with automatic detection. It performs especially well in English, German, French, Spanish and many more.

Which file formats are supported?

All common audio and video formats: MP3, WAV, M4A, FLAC, OGG, MP4, WebM, MOV, AVI, MKV and many more. Files up to 10 GB are supported.

Can I reset my password?

Yes, but because your encryption key is derived from your password, a password reset means existing transcripts can no longer be decrypted. That's the price of true client-side encryption.

How are your AI models trained?

Our models are trained and refined exclusively with paid speakers and open sources such as Mozilla Common Voice. Your data is encrypted in your browser and stays encrypted on our servers.

Ready for secure transcription?

Get started now – encrypted from the first second.

Transcribe audio with accurate AI transcription | scryp