Speak instead of type. Listen instead of read. Expand a shortcut into a paragraph. Edit text by voice. Everything runs locally — $10 once, no subscription.
Every model runs locally on your hardware. No server, no account, nothing ever sent.
Zero cloud. Audio, text, and commands are processed by open-source models on your hardware. Nothing leaves your machine.
Local models mean your words appear in real time — no round-trips, no buffering. Works on a plane, underground, with no internet at all.
Parakeet for real-time streaming in 25 languages. Whisper for depth across 99. Switch any time.
Global hotkeys for dictation, read-aloud, AI commands, and snippet expansion. Works system-wide in any app — fully customisable.
Select any text and press a hotkey — Dictar reads it back with 51 natural-sounding local voices. Articles, emails, docs. No cloud TTS, all offline.
Type /greet and press Space — it expands into your full greeting. Supports {date}, {fill}, and AI-drafted text. All processed locally.
Select any text, say "make this professional" — an on-device AI rewrites it in seconds. No API key, no upload, no account needed.
No account. No config. No learning curve.
Pay $10, get a license key by email. Paste it in — done. No account, no setup.
75 MB for speed or 3 GB for accuracy. Stored locally. Switch any time.
Alt+Space to dictate, separate keys for TTS and commands. Link modes to apps.
Dictate anywhere, read anything back, expand shortcuts. Your keyboard becomes optional.
"I compose first drafts by speaking, then tighten them with voice commands. My output is 3× what it was — and nothing ever leaves my machine."
"The snippet triggers alone were worth the price. /proposal in Slack, done. /followup in Gmail, done. Email time cut in half."
"Privacy was the deciding factor. My clients' work stays on my machine — no cloud, no uploads, no exceptions."
No subscription. No cloud. No catch.