Hold to Talk, Release to Paste
Your voice becomes ready-to-use written text in 3 seconds
Free tiers from Groq & Gemini cover daily use — transparent, pay-as-you-go pricing
Actively maintained — see latest improvements & features
Notch-style Interactive HUD
A MacBook-notch-inspired status indicator stays pinned at the top of your screen without blocking your work. From recording to transcribing to done, you always know what SayIt is doing.
Open Source (MIT)
Full source code on GitHub under the MIT license — free to use, modify, and commercialize. No vendor lock-in, community-driven evolution.
View SourceVoice Input in 3 Steps
Fn on macOS, Right Alt on Windows
Whisper transcribes, LLM refines into written text
Text appears right at your cursor
Why SayIt?
From releasing the key to seeing the text — faster than switching input methods
Both Groq and Gemini offer daily free quotas — zero cost for regular use
Recordings and history stay on your computer, never uploaded to any cloud
Slack, VS Code, Email, GitHub — just hold and speak
Get your own Groq or Google Gemini API Key — the free daily quota is enough for regular use
Even beyond the free tier, one hour of transcription costs about $0.10 USD
You pay the API provider directly (Groq / Gemini / OpenAI / Anthropic) — no middleman, no subscription
Compare: paid services like Typeless start at $10+/month. SayIt lets you pay only for what you use — and most of the time, it's free
Get Your Groq API Key in 3 Minutes
Go to the API Keys page, click "Create API Key", enter a name, and create
Go to API Keys PageCopy the key, open SayIt settings, paste it in — done!
Features
System-wide Hotkey
Hold Fn in any app to record — no window switching needed
Whisper Fast Transcription
Multiple API providers supported — get your transcript as fast as 1 second after releasing the key
AI Speech-to-Writing
Automatically removes filler words and refines into clean, usable sentences
Custom Dictionary
Add your own terms so "Fortuna" stays "Fortuna" instead of being misheard
Local Privacy Storage
Recordings and text stay in your local SQLite — nothing uploaded to any cloud server
Cross-Platform
Native apps for macOS + Windows — setup takes 2 minutes
Voice-Edit Selected Text
Select text, hold the hotkey, give voice instructions (translate, rewrite, summarize) — AI replaces the original
Recording History & Playback
Every recording is auto-saved — replay audio, review transcripts, or re-transcribe anytime
Sound Feedback & Instant Cancel
Audio cues for recording start, stop, and errors; press ESC anytime to cancel recording or AI processing
Rich Customization
macOS: Fn / Option / Control / Command / Shift. Windows: Right Alt / Left Alt / Control / Shift. Custom key recording supported. Combo hotkeys: double-tap or long-press to switch modes
Hold mode: hold to record, release to stop. Toggle mode: press once to start, press again to stop — great for long dictation
4 AI providers: Groq (free), Google Gemini (free tier), OpenAI (recommended), Anthropic. Multiple models per provider. Whisper: Large V3 or V3 Turbo
Three modes: Minimal (fix typos, remove fillers), Active (restructure with bullet points), Custom (fully custom prompt). Defaults optimized for clean written text
Below a set word count, AI refinement is skipped — raw transcript is pasted directly to avoid rewriting short commands
System speakers auto-mute when recording starts, auto-restore when done — prevents notification sounds from being captured
Choose your preferred microphone in settings with live volume preview — no longer limited to the system default device
Who Is It For?
PR descriptions, code review replies, technical docs — speaking is 3x faster than typing
Dictate meeting summaries in 20 seconds — auto-formatted into clear bullet points
Email replies, Slack messages, reports — your voice is more efficient than your fingers