Hold to talk.
Paste anywhere.
Native macOS menu-bar app for on-device dictation in 99+ languages, plus a Claude assistant that sees your active window. Hold a key, speak, release. Text appears where your cursor is. €29 once, every Mac, lifetime updates.
One payment. Forever yours.
€29 once. Lifetime updates included. No subscription, no tiers, no upsells. Ever.
Nothing reaches our server.
No telemetry, no analytics, no account, no tracking. Signocore cannot see what you transcribe, ask, or paste. None of it ever touches our servers.
No license to activate.
Buy, download, run. No license key, no account, no activation flow. It works the moment you launch it.
What hold-to-talk dictation actually does.
Hold-to-talk dictation converts speech to text on your Mac without launching a separate app or copying anything between windows. You hold a global hotkey, speak naturally, and release. The transcribed text appears wherever your cursor is: editor, browser, mail, chat, anywhere.
Faster than built-in macOS dictation for most workflows, because there is no UI to navigate, no language picker to click through, and no copy-paste between a transcription pane and the destination. The hotkey is the entire interface.
Edicta runs the transcription on-device using OpenAI Whisper through WhisperKit, running on Apple Silicon's Neural Engine. Audio never leaves your Mac. No cloud round-trip, no audio uploaded to any server, no internet required after the model is downloaded once.
The same hold-to-talk pattern drives Edicta's Claude AI assistant: hold a different hotkey, ask a question, and Edicta sends your spoken question along with a screenshot of your active window to Claude. The streamed answer is pasted at your cursor as it generates.
Hold a key. Speak. Release.
No window switching, no copy-paste shuffle. The result appears wherever your text cursor is.
Dictation
Speech to text, on-device.
Hold the dictation hotkey, speak, release. WhisperKit transcribes locally on your Mac. No audio ever leaves the device. Text gets pasted at your cursor.
- · 99+ languages including Danish, Swedish, Norwegian, Finnish
- · On-device via OpenAI Whisper (no internet required)
- · Works in any text field: editor, browser, mail, chat
- · Floating waveform shows recording status
Assistant
Claude understands your screen.
Hold the assistant hotkey, ask a question, release. Edicta captures your active window, sends it with your spoken question to Claude, and pastes the streamed answer at your cursor.
- · Active window screenshot for visual context
- · Streamed response: words appear as they are generated
- · Powered by Claude (Anthropic API)
- · Bring your own API key, pay only for what you use
No dock icon. No window taking up space.
A small icon in your menu bar. The hotkey is the interface.
Quick controls
Switch transcription language, see hotkeys, and access settings, all from the menu bar.
Live feedback
A floating pill with animated waveform shows exactly when Edicta is listening, transcribing, and pasting.
Native settings
Standard macOS settings window with tabs for general, hotkeys, languages, Claude, permissions, and about.
Your voice stays on your Mac.
Dictation runs entirely on-device. WhisperKit uses Apple Silicon's Neural Engine to transcribe locally. Audio is never uploaded, stored, or transmitted. Not to Signocore, not to OpenAI, not to anyone.
Assistant uses the Anthropic API directly. Your active-window screenshot and spoken question go from your Mac straight to Claude using your own API key. Signocore never sees your data, prompts, or responses.
No telemetry, no analytics, no accounts. Edicta doesn't phone home. No tracking, no usage logs, no sign-up. We don't even offer accounts.
Dictate in any language Whisper supports.
Switch transcription language from the menu-bar dropdown in one click.
Plus 74 more, from Afrikaans to Welsh. Every language Whisper supports.
Built native. No subscription. No bloat.
Rebindable hotkeys
Pick any global key combination for dictation and assistant, including modifier-only keys like Right Option.
Active-window context
Assistant captures only your active window. Not the whole screen.
Per-session language
Switch transcription language from the menu bar without restarting.
Pasteboard-safe
Snapshots your clipboard before pasting, restores it after. Never lose what you copied.
Launch at login
Starts when you sign in, ready in your menu bar before you reach the keyboard.
Native Swift, no Electron
Built in Swift 6 with SwiftUI. Small footprint, fast launch, low memory.
No telemetry
Edicta doesn't collect or transmit any usage data. Nothing to opt out of.
Lifetime updates
One purchase. Every future version included. No subscription, no upgrade fees.
Sound feedback
Optional chimes when recording starts and stops, so you don't have to look at the screen.
What you need to run Edicta.
One price. Yours forever.
No subscription. No license activation. No feature tiers. Buy once, download, use it on every Mac you own.
€29 once
Lifetime updates included.
- · All features included, no Pro tier
- · Use on every Mac you own
- · Lifetime updates, no subscription
- · Pay once with credit card via Stripe
- · Email me directly if anything's wrong
Apple Silicon Mac required. Anthropic API key needed for assistant mode.
Common questions about Edicta.
Why is Edicta not on the App Store?
+
Edicta uses global hotkeys, screen recording, and pasteboard access. App Store sandboxing doesn't allow those. Distributing outside the App Store means signed and notarised by Apple, downloaded directly from Signocore.
Does dictation work offline?
+
Yes. WhisperKit runs entirely on your Mac using the Neural Engine. Once the Whisper model is downloaded on first launch, dictation works without an internet connection.
What data do you collect?
+
Nothing. Edicta has no analytics, no telemetry, and no account system. Your purchase email is stored only to send you the download link.
Which Macs are supported?
+
Edicta requires macOS 26 (Tahoe) or newer running on Apple Silicon (M1, M2, M3, or M4). Intel Macs are not supported because WhisperKit relies on the Neural Engine for fast, on-device transcription.
How do updates work?
+
Edicta uses Sparkle to deliver updates automatically. New versions install themselves with one click. Every future update is included in your purchase. Forever, no subscription.
Do I need an Anthropic API key?
+
Only for the Assistant feature. You bring your own key from platform.anthropic.com so you pay only for what you use, and your prompts never go through a Signocore server. Dictation works without any key.
What if it doesn't work for me?
+
Edicta is sold as-is at a low one-time price, so I don't offer a scheduled refund window. But I want happy customers. If something's broken, won't run on your Mac, or behaves differently than the page describes, email me and we'll sort it out.