Uploading private recordings to someone else's cloud
Interviews, therapy sessions, legal calls - sent to a server you don't control, kept for who knows how long.
Dictea records and transcribes audio and video right on your computer - with automatic speaker detection and a built-in step-by-step documentation tool. No cloud, no account, no subscription. Unlimited and offline.
No account · No credit card · Runs locally
Me: Shall we start with the quarterly review?
Guest: Sure - I've got the slides ready to share.
Me: Great, let's walk through the first one.
Dictea is a desktop app that records and transcribes audio and video 100% locally on your own computer. Whisper-class transcription runs offline with automatic speaker detection, plus a built-in step-by-step documentation tool. There is no cloud, no account and no subscription, and it works on macOS and Windows.
Transcription should make your life easier - not put your most private conversations at risk.
Interviews, therapy sessions, legal calls - sent to a server you don't control, kept for who knows how long.
Subscriptions that never end and per-minute fees that quietly add up the more you actually use them.
Raw transcripts full of filler, wrong breaks and formatting you have to fix by hand before anyone can read them.
A wall of text with no idea who spoke when - so you scrub the audio back and forth just to label two voices.
Here's the shift.
Everything runs on your computer. Nothing gets uploaded.
Draft - operator to approve wording & facts
Why Dictea exists
Hi, I'm Mathias. I build Dictea on my own, from the island of Madeira.
I started it for a simple reason: I record real conversations - interviews and personal talks - and I never felt comfortable handing them to someone else's cloud. So I built the tool I actually wanted to use: everything runs on your own computer, and your recordings simply stay yours.
There's no big company behind this - just one developer who'd rather get privacy right than cut corners.
- Mathias Kusch, maker of Dictea
Capture microphone and system audio on separate tracks, or drop in any existing audio or video file.
Whisper-class models run on your computer and turn speech into text with automatic speaker detection - offline.
Rename speakers, fix segments, then export to any of 9 formats - or a ready-made step-by-step document.
Record, transcribe, edit and document in one tool - and none of it leaves your computer.
Microphone and system audio captured separately in an OBS-style mixer with gain, mute and solo. The per-source WAVs are kept.
Per-segment speaker detection; with multiple tracks it's channel-based (Me vs. Guest). License-free and fully local.
An editable transcript editor: rename or merge speakers, delete and reorder blocks, inline-edit with a synced player.
Every click captures a screenshot and a numbered step marker, with voice narration - a finished doc as DOCX, MD, HTML or PDF.
TXT, MD, SRT, VTT, JSON, segmented JSON, DOCX, CSV and PDF - ready for subtitles, docs or further processing.
Audio and transcripts stay local. No cloud API, no tokens or per-minute fees, telemetry off by default.
No account · No credit card · Runs locally
One architectural choice - everything runs on your computer - pays off five different ways.
From microphone to finished transcript, every step runs inside your computer. No upload, no server, no API sits in between.
Works even with Wi-Fi switched off.
Two very different paths
Microphone → your computer → transcript
Your audio never leaves the device.
Microphone → upload → their servers → back
Your audio leaves your machine.
No invented reviews, no fake user counts - just honest signals you can check yourself.
Works with Wi-Fi off
Pull the network cable and Dictea keeps transcribing. That's the proof it's local.
No account required
Download, start the trial, transcribe. There's nothing to sign up for.
No telemetry
Usage tracking is off by default. We don't watch how you work.
Buy once, no lock-in
Choose Lifetime and there's no subscription holding your files hostage.
30-day money-back
If it's not for you, get a full refund within 30 days. No hard feelings.
Dictea runs modern Whisper-class speech models directly on your computer. You get the transcription quality you'd expect from the cloud - without anything leaving your device.
Get Dictea
Install it, hit record, and transcribe locally. The 7-day Pro trial runs inside the app - no account, no credit card.
macOS 12+ · Windows 10/11 (64-bit)
Pick a subscription or buy it once. Handled by creem.io as Merchant of Record - prices in EUR incl. applicable tax.
Try the basics
No account · No credit card
Full Pro features · No credit card · Cancel anytime
Full Pro features · No credit card · Cancel anytime
No subscription, ever. Stays on your computer.
1 device
Up to 3 devices
Not a feature list for its own sake - these are the things you walk away able to do.
Replace a typical cloud subscription - often €15-30 a month, forever - with a single one-time purchase.
Typical cloud tool
€15-30 / month, indefinitely
Dictea Lifetime
Pay once, keep it
30-day money-back guarantee
Try Dictea risk-free. If it isn't for you, ask for a full refund within 30 days - no questions, no hard feelings.
No friction, no surprises - and your data stays on your computer the whole time.
As soon as checkout is done, you receive your personal license key.
Download Dictea for your platform, enter your key once, and Pro unlocks instantly.
Record or drop in a file and transcribe - fully local, with nothing uploaded.
Operator TODO: exact key delivery (e.g. by email) is not finalized yet - confirm the flow before launch.
No account · No credit card · Runs locally
Yes. All audio, transcription, speaker detection, editing and export happen on your computer. There is no cloud transcription API. The only thing that ever leaves your machine is license data (see the privacy policy).
Download Dictea and transcribe your first file in minutes - fully local, nothing uploaded.
No account · No credit card · Runs locally