Your voice.
Every app.
Dictate text, search any engine by voice, and execute actions across Slack, Linear, and more — all from your keyboard. No typing. No switching windows.
Private beta · Limited spots
Three modes. One key.
Hold fn and speak. mrmr figures out the rest.
Talk anywhere.
Text appears.
Hold fn and speak. mrmr transcribes, removes filler words, adds punctuation, and pastes clean text wherever your cursor is. Works in every text field on your Mac.
- System-wide — works in any app with a text field
- Smart formatting strips filler words and adds punctuation
- Hands-free mode: double-tap fn, press fn + space, or hit space mid-dictation
- Clipboard fallback if focus is lost
Two ways
to search.
mrmr gives you a quick-search shortcut and an intent-based search that works while you dictate. Both support custom search engines.
Searches your default search engine. Press s again to cycle through other engines — Google, Perplexity, Claude, ChatGPT, or any custom engine you've added. Speak, release, results open.
Just dictate normally. Say "reddit voice dictation macos" and mrmr opens Reddit with that search. Say "youtube how to setup claude code" and YouTube opens. If it's not a search, text inserts at your cursor as usual.
Add your own search engines with aliases and a URL template. Works with both quick search and intent-based routing.
Speak commands. Apps execute.
Hold fn + shift and tell mrmr what to do. mrmr parses your intent, chains the actions, and executes them across your apps — with confirmation before anything runs.
Chain actions — one command triggers multiple apps 
Confirmation UI — nothing runs without your approval 
Attach screenshots, GIFs, and more for context 
Slack, Linear, Google Calendar, Google Meet, Zoom — more coming soon Works with your tools
Dictation works in any text field. Search and actions work across these.
Slack 
Notion 
Linear 
WhatsApp 
Cursor 
LinkedIn 
X 
ChatGPT 
Claude 
Teams 
Obsidian Slack Notion Linear WhatsApp Cursor LinkedIn X ChatGPT Claude Teams Obsidian Slack Notion Linear WhatsApp Cursor LinkedIn X ChatGPT Claude Teams Obsidian Slack Notion Linear WhatsApp Cursor LinkedIn X ChatGPT Claude Teams Obsidian 
Google 
Perplexity Claude ChatGPT 
YouTube 
Reddit + your own Google Perplexity Claude ChatGPT YouTube Reddit + your own Google Perplexity Claude ChatGPT YouTube Reddit + your own Google Perplexity Claude ChatGPT YouTube Reddit + your own Slack Linear 
Google Calendar 
Google Tasks 
Google Meet 
Zoom 
GitHub Notion 
Jira Microsoft Teams 
Gmail 
Microsoft Outlook Slack Linear Google Calendar Google Tasks Google Meet Zoom GitHub Notion Jira Microsoft Teams Gmail Microsoft Outlook Slack Linear Google Calendar Google Tasks Google Meet Zoom GitHub Notion Jira Microsoft Teams Gmail Microsoft Outlook Slack Linear Google Calendar Google Tasks Google Meet Zoom GitHub Notion Jira Microsoft Teams Gmail Microsoft Outlook Get early access
mrmr is in private beta. Join the waitlist or book a demo call for fast-track access.
FAQ
Dictation works system-wide in any text field on macOS. Action Mode currently supports Slack, Linear, Google Calendar, Google Meet, and Zoom, with GitHub, Notion, and Jira coming soon.
macOS dictation gives you raw speech-to-text. Siri handles basic system tasks. mrmr does both — plus smart formatting, intent-based search routing, and cross-app action execution. It's a voice interface layer on top of your entire workflow.
Yes. Audio is processed securely and immediately deleted. We never store recordings or use them for training. Only the transcribed text remains on your device.
mrmr is in private beta. Join the waitlist and we'll send you access as spots open up. Want to skip the line? Book a 15-minute demo call for immediate access.
mrmr parses your voice command, shows you exactly what it will do, and waits for confirmation. Nothing executes without your approval. You can review, edit, or cancel any action before it runs.
No, mrmr requires an internet connection for transcription and action execution. Your audio stays on your device — we don't store recordings.
Text typically appears within 1–2 seconds of releasing the fn key. mrmr uses optimized transcription infrastructure designed for low latency — fast enough to feel like a natural extension of typing.
Two ways. fn + s opens a quick search on your default engine — press s again to cycle through others. Or just dictate normally with fn — if mrmr detects a search intent (like "youtube how to setup claude code"), it opens the right site automatically. Both methods work with custom search engines you add yourself.
Yes. mrmr uses OAuth for all integrations — we never see or store your passwords. Permissions are scoped to only what's needed, and you can revoke access at any time from your account settings in each app.
mrmr needs Accessibility permission (to insert text at your cursor) and Microphone access (to capture your voice). All standard macOS permission prompts — nothing runs in the background without your knowledge.
mrmr currently supports English. Multi-language support is on the roadmap — if you need a specific language, let us know and we'll prioritize accordingly.
Yes. During the private beta, mrmr is free to use. We'll announce pricing before any paid plans launch, with plenty of notice.
Stay in the loop
Get product updates and tips for voice-first workflows. No spam.
