Type with your
voice.
Hold a key. Talk. Your words appear wherever your cursor is. And when you're in a meeting, Parley records it, tells the speakers apart, and writes the notes. Nothing leaves your computer.
Free · No account · No signup · Works offline
I've finished testing the new dictation model. Ready for your review whenever you're free.
Awesome work. Can you share a quick summary in Slack so I can read it before our meeting tomorrow?
How it works
Three steps. That's it.
Hold your key
Press and hold Right Ctrl (Windows) or Right ⌘ (Mac) — or pick any key you like on first launch (no Right Ctrl? Copilot-key and Korean/JIS keyboards are covered).
Just talk
Speak naturally. Say exactly what you'd normally type out by hand.
Let go
Release the key. Your words are typed out and pasted right where your cursor is.
And for meetings
Click the bird
A little cockatiel sits in the corner of your screen. Click him and choose Record meeting — or Just record me for voice memos.
Have your meeting
Parley captures you and the other side, transcribing live and saving to disk as it goes.
Click Stop
The transcript and AI notes open in your browser, speakers labeled and ready to rename.
He also listens while you dictate — and cheers when your text lands.
Features
Built for people who type a lot
Whether you're writing emails, messages, docs, or code comments — speaking is faster than typing.
Sits in your menu bar / system tray
No app to open. No window to find. Just Parley lives quietly in your menu bar (Mac) or system tray (Windows), always one hotkey away. It stays out of your way until you need it.
Sits in your menu bar / system tray
Completely offline
Your voice never leaves your computer. No cloud processing, no accounts, no data collection. The speech model runs entirely on your machine.
Completely offline
Data sent to the cloud
Five languages, auto-detect
English, Chinese, Japanese, Korean, and Cantonese. Pick one or let the app detect the language automatically as you speak.
Works everywhere
Any text field, any app. Emails, Slack messages, Google Docs, code editors, search bars — if you can type in it, Just Parley can type in it for you.
New
Your meetings, written down — privately
Parley records both sides of any call, transcribes it live, and hands you the notes. No bot joins your meeting, and no audio ever leaves your computer.
Record any call
Captures your mic and the other participants (system audio) in any app — Zoom, Meet, Teams, or a call in your browser. Requires macOS 14.4+ or Windows.
Who-said-what
Offline speaker identification labels the other side Speaker 1, 2, 3 as they talk. Click a name on the review page to rename them.
AI notes without the cloud
A local model (via the free Ollama app, optional) turns the transcript into a summary, decisions, and action items. Skip it and you still get the full transcript.
A tidy review page
Every meeting gets a page in your browser: audio, colored speakers, timestamps, and a searchable index. No account, no upload — it's all files on your disk.
“I work with a team of AI agents every day — ChatGPT, Claude, Cursor. Typing was becoming the weak link. I was spending more time on prompts than on actual problems, and I needed to change the way I was working. The dictation tools out there either sent my audio to the cloud, charged monthly, or both. I wanted something that ran entirely on my machine and just worked. So I built it — and just like the research says, my productivity went up 3x.”
Compare
How we stack up
Paid dictation apps compared. Only the features that matter.
Just Parley | Superwhisper | Wispr Flow | MacWhisper | Voibe | |
|---|---|---|---|---|---|
| Price | Free | $8.49/mo | $15/mo | $79.99 once | $49.99 once |
| Offline | ✓ | ✓ | ✗ Cloud | ✓ | ✓ |
| Works in all apps | ✓ | ✓ | ✓ | ✗ Transcription only | ✓ |
| English | ✓ | ✓ | ✓ | ✓ | ✓ |
| 中文 Chinese | ✓ | Weak | ✓ | ✓ | ✗ |
| 日本語 Japanese | ✓ | ✓ | ✓ | ✓ | ✗ |
| 한국어 Korean | ✓ | ✓ | ✓ | ✓ | ✗ |
| 粵語 Cantonese | ✓ | ✗ | ✗ | ✗ | ✗ |
| Meeting recording | ✓ Mic + system audio | ✗ | ✗ | Files only | ✗ |
| Speaker identification | ✓ On-device | ✗ | ✗ | Pro | ✗ |
| AI meeting notes | ✓ Local (Ollama) | ✗ | ✗ | Cloud API | ✗ |
| Privacy | 100% on-device | On-device | Cloud processed | On-device | On-device |
What about Apple's built-in Dictation?
Apple Dictation is free and works well for basic use — but it has limits. It doesn't work in Electron apps like Slack, Discord, or VS Code. On Intel Macs, your audio is sent to Apple's servers. And accuracy for Chinese, Japanese, and Cantonese varies significantly. If you dictate in English occasionally, Apple's option is fine. If you type a lot, work across multiple languages, or care about privacy on every Mac, Just Parley fills the gap.
Works in your favourite apps
Just Parley works everywhere — chat, code, notes, design, and more.
Chat & Messaging
FAQ
Common questions
What languages are supported?
English, Chinese (Mandarin), Japanese, Korean, and Cantonese. You can pick one or set it to auto-detect.
Does it need an internet connection?
No. The speech recognition model runs entirely on your computer. Your audio is never sent anywhere — not to us, not to anyone.
Which devices are supported?
macOS 13 (Ventura) or later (Apple Silicon and Intel). Windows 10 or later.
Is it a subscription?
Just Parley is completely free. No subscription, no signup, no card. Donations are welcome if you want to support development, but they're never required.
How accurate is it?
Just Parley uses SenseVoice, a state-of-the-art speech recognition model. It handles natural speech, punctuation, and multiple languages well — and it's getting better with each update.
Can I change the hotkey?
Yes — the app asks you to pick your key on first launch (default: Right ⌘ on Mac, Right Ctrl on Windows). Caps Lock, modifier keys, F1–F12, and even the 한자/変換 keys on Korean and Japanese keyboards all work.
Can it record my meetings?
Yes. Parley records your microphone and the other participants (system audio) in any meeting app, transcribes live, and writes notes when you stop — all offline. On macOS 14.4+ it asks once for the System Audio Recording permission.
Is speaker identification private?
Yes. Voices are clustered entirely on your device using small local models (~40 MB, downloaded once). Speakers are labeled Speaker 1, 2, 3 and you can rename them on the review page. No voiceprints leave your computer.
Do I need Ollama?
Only for AI meeting notes, and it's optional. Ollama is a free app that runs AI models locally — with it, Parley writes a summary, decisions, and action items. Without it, you still get the full transcript.