One keystroke. No clicks.
Type like you text. Select it, press your instant-polish shortcut, and it's rewritten in place — right inside the app you're already in.
Private by design. Fast by default.
No accounts, no API keys, no cloud round-trips. Just better writing, instantly.
Runs entirely on your Mac
Zero internet. Your text never leaves the machine — nothing is sent to any server, ever. True on-device privacy.
Powered by Apple Intelligence
Uses Apple's on-device foundation model on your Mac's own silicon. No API keys, no subscriptions, no cost.
Four tones, one keystroke
Fix Grammar, Professional, Friendly, or Concise. Preview and tweak, or replace instantly without lifting your hands.
Works in every app
Slack, Mail, Notes, your browser — anywhere you can select text. Lives quietly in the menu bar, no Dock clutter.
Select text. Press a key. Done.
Prefer different keys? Both shortcuts are fully customizable — set your own from the menu bar under Shortcuts…
Any app. Zero internet.
TextPolisher works anywhere you can select text — Slack, Mail, Notes, your browser, your IDE — and runs entirely on-device. No network, no API calls, no telemetry. 0 bytes ever leave your Mac, so it works on a plane, offline, or anywhere you don't trust the network.
One price. Everything included.
No tiers, no feature gates, no surprises.
- All four tones — Fix Grammar, Professional, Friendly, Concise
- Instant replace & preview overlay
- Customizable keyboard shortcuts
- Works in every app — Slack, Mail, browser, IDE
- 100% on-device — 0 bytes ever leave your Mac
- No usage limits, no API costs
- Free updates
Up and running in a minute.
Enable Apple Intelligence
System Settings → Apple Intelligence & Siri → turn it on.
Drag to Applications
Open the .dmg and drop TextPolisher into your Applications folder.
Open it
Launch it — look for the wand in the menu bar. No Dock icon, no window.
Grant Accessibility
Allow it in Privacy & Security → Accessibility so it can read and replace your selection.
No Apple Intelligence on your Mac? It falls back to a local Ollama model — still fully on-device.