Don't type,
just dictate.
Private, on-device speech-to-text. $10 once, yours forever.
One-time payment · No subscription · Every future update free
Pay once,
own it forever.
No monthly fees, no per-minute charges, no cloud costs. Every word you speak is processed entirely on your machine. Your voice never leaves your device.
One-time purchase (normally $15). No subscriptions, no hidden fees. Ever.
Languages supported. Whisper for breadth, Parakeet for speed. All running locally.
Transcription latency. Speak naturally, see your words appear in real-time.
Everything you need to replace
your keyboard — nothing you don't.
Other dictation tools charge $10/month and send your voice to the cloud. Dictar costs $10 once — and it's yours for life.
DictarPrivate by design
Your voice never leaves your machine. No cloud, no uploads, no third-party processing. Unlike subscription tools, there's no server to get hacked.
One hotkey, instant paste
Hold Alt+Space to record, release to transcribe. Your words are automatically pasted wherever your cursor is — any app, any text field.
99 languages
Choose Parakeet for fast streaming in 25 languages, or Whisper for coverage across 99 languages. All models download once and run locally.
Live transcription
See your words appear in real-time as you speak. A floating overlay shows a waveform visualization and streaming text so you know exactly what's being captured.
On-device AI models
Powered by Parakeet and Whisper — the same models used by researchers and professionals worldwide. Download once, use forever, no internet required.
$10. That's it.
Launch price — normally $15. Built on open-source models, honestly priced. Your $10 funds the app, the UX, and every feature on our roadmap. No subscriptions, no "premium tiers".
Up and running in under a minute.
No accounts, no configuration, no learning curve. Buy it, download it, start talking.
Pay $10, get a license key instantly via email. Paste it into the app — done in 10 seconds.
Choose a speech-to-text model. Smaller ones are faster, larger ones are more accurate. Download once.
Hit Alt+Space to start recording. A floating overlay appears with a live waveform and streaming text.
Release the hotkey. Your words are automatically pasted wherever your cursor is — any app, any text field.
Questions & answers
The models are open source — why pay?
Yes, Dictar is built on open-source models (Whisper, Parakeet) and we're upfront about that. What you're paying for is the app itself — the hotkey system, live overlay, auto-paste, model management, cross-platform builds, and every new feature on our roadmap. The $10 funds active development so we can keep shipping features like read-aloud, voice commands, and more. No cloud costs, no subscriptions — just honest software.
What do I get for $10?
A polished app that just works, plus every future feature and update at no extra cost. This is a launch offer — the price will go up to $15. One license key for one device — you can deactivate and move it to a different device anytime.
Is my audio sent to any server?
No. Dictar processes everything locally using downloaded AI models. No audio data ever leaves your device — not even anonymized data. There are literally no servers involved.
What models does Dictar use?
Two model families: Parakeet (by NVIDIA) for fast streaming in 25 languages, and Whisper (by OpenAI) for 99 languages. Models range from 75MB to 3GB. All run locally on your machine.
Does it work offline?
Completely. Once you download a model, Dictar never needs the internet again. Take it on a plane, use it in a cabin — it just works.
Can I use it on multiple devices?
Each license activates on one device at a time. Need to switch? Just deactivate from the old device and activate on the new one — takes two clicks.
macOS says "Dictar is damaged"?
Standard macOS security warning for apps outside the App Store. Open Terminal, run: xattr -cr /Applications/Dictar.app, and open Dictar normally. One-time fix.
What are the system requirements?
macOS or Windows with a modern CPU. No dedicated GPU required. The app is lightweight — most of the size comes from whichever AI model you choose to download.
See what's coming next.
Read-aloud, voice commands, speaker diarization, and more. Check out our roadmap to see what we're building.
View roadmapStop paying monthly
for your own voice.
$10 once. No subscription. No cloud. No limits. Works on macOS and Windows with free updates forever.