Post messages, tag people, drop links into any channel.
Voice to action
Hold a key and talk. mrmr's voice agent listens, talks back, and takes action across your apps, the web, and your Mac, confirming with you before anything runs.
Free during private beta.
A real back-and-forth, not one-shot commands. It asks when it needs to.
Reads are free; anything that writes is confirmed first.
Built for macOS. Lives in the menu bar, one keystroke away.
Mic opens only when you trigger it. No always-on listening, no training.
Works with the tools your team already uses.
Agent Mode connects to your work apps, plus built-in web search and on-Mac file search. Dictation works anywhere you can type.
Create issues, set priority, assign owners, link PRs.
Schedule, move, or cancel events with a sentence.
Capture and check off tasks without breaking flow.
Spin up a Meet and send the link where you need it.
Generate a Zoom link and share it across apps.
Capture notes and drop them into the right page.
Draft and send replies from anywhere on your Mac.
Check availability and manage bookings by voice.
Pull up scheduled meetings and share your links.
Look up contacts and update your CRM as you talk.
Set, update, and complete reminders mid-conversation, hands-free.
Open issues and PRs, comment, request review.
Live answers with cited sources, right in the conversation.
Finds files on your Mac and opens documents and folders.
Searches your browser history and bookmarks and opens the right link.
Copies results to your clipboard and opens the right link directly.
Delegate bigger tasks to an agent that works while you keep talking.
Update deals, log notes, and look up contacts by voice.
Create tickets and update fields by voice.
Draft messages and schedule sends in your inbox.
Send messages and start calls hands-free.
Dictation Mode works in any app with a text field.
Just talk.
Your apps get it done.
Press fn + shift and speak naturally. The agent listens, asks clarifying questions, chains the steps across your apps, and reads results back to you. Anything that sends, creates, or changes is confirmed before it runs. Why speech-to-action matters →
Animated demo of the Agent Mode overlay: a floating status pill cycles through Listening, Thinking, and Speaking while the agent checks a calendar, moves a meeting and posts to Slack, summarizes a channel, adds an Apple Reminder, runs a workflow, and delegates research to a background agent. When a task is delegated, a small sub-agent window pops into the top-right corner, collapses to an icon while it works, and re-expands with the finished result.
“Create an urgent ticket for the authentication issue on the login screen, assign it to the dev team and Hamza, message #general that we're tracking the fix and it'll be out soon, tag Hamza, and schedule a follow-up at end of day.”
- 01 Create Linear issue “Authentication issue on login screen”
- 02 Set priority to Urgent
- 03 Assign to Dev team and Hamza
- 04 Post in #general, @Hamza tagged
- 05 Schedule follow-up today at 5:00 PM
Have a conversation
Not rigid commands. Follow up, correct it, ask what happened, all mid-task.
Confirm before it acts
See the apps, fields, recipients, and content before anything is sent or created.
Delegate the heavy stuff
Hand bigger tasks to a background agent and keep talking. Ask for the result anytime.
Beyond your apps
Web search with sources, browser history, files, clipboard. All built in.
Built for the work you repeat every day.
Four patterns mrmr handles well on day one, and the kind of work most beta users automate first.
“Create a bug ticket and message the team with the link.”
mrmr creates the Linear issue, copies the link, and posts the Slack update.
“Move my 3pm meeting with Sarah to 4pm.”
mrmr finds the event, prepares the change, and updates the invite after confirmation.
“Tell #product I'll push the release notes by end of day.”
mrmr drafts the Slack message and sends it after approval.
“What's the latest on the OpenAI API pricing?”
mrmr searches the web and reads back a live, citation-backed answer without leaving the conversation.
“Find the Q3 deck and remind me to send it at 4pm.”
mrmr finds the file, opens it, and creates the Apple Reminder.
Dictation rides along.
Agent Mode is the core. Dictation is the second mode: fast, polished voice typing anywhere on your Mac.
Speak. Polished text appears.
- Hold fn and speak into any text field
- Cleans up filler and adds punctuation
- Hands-free with fn + space
- Works system-wide in every app
Animated demo of the Dictation Mode overlay: a recording pill with a live waveform and timer streams your words as you speak, fillers included, then mrmr cleans them up and inserts the polished sentence into the active app, cycling through Gmail, Slack, and Apple Notes examples.
Not just dictation. Not just shortcuts.
| Product type | What it does | Limitation |
|---|---|---|
| macOS Dictation | Turns speech into text | Does not execute workflows |
| AI dictation apps | Better voice typing | Mostly still text output |
| Siri / Shortcuts | Basic commands and automations | Brittle for real work-app workflows |
| mrmr | A voice agent that takes confirmed action across your apps and Mac | Mac-only, invite required |
Private beta is best for people who already live in their work apps.
- You use a Mac every day
- You work in Slack, Linear, and Google Calendar
- You repeat small workflow tasks throughout the day
- You are comfortable trying a private beta
- You want direct access to the founder
- You need Windows support
- You need enterprise admin controls today
- You need offline-only transcription
- You mainly want long-form meeting transcription
- You need every integration to be production-ready now
Get private beta access.
Hi, I'm Hamza, the founder. mrmr is invite-only right now. I'm onboarding every beta user personally so I can see which workflows matter, where the product breaks, and what to build next.
- Early access to Agent Mode
- Direct setup help
- Founder pricing when paid plans launch
- Priority input on new integrations
- Direct feedback channel with the founder
Questions, answered.
mrmr is a voice-first Mac app with two modes: a voice agent that takes confirmed actions across your work apps, the web, and your Mac, plus system-wide dictation that turns speech into polished text.
Agent Mode is a continuous voice agent. Press fn + shift and just talk. It listens, talks back, and takes action across your connected apps (like creating a Linear ticket and posting the link in Slack), searches the web, and works with files and reminders on your Mac. It confirms before it sends, creates, or changes anything.
No. Reads and searches happen freely, but anything that writes to your apps or your Mac is shown to you first. You can edit the details, and nothing runs until you approve.
Hold fn, speak, release, and polished text appears at your cursor in any app. It cleans up filler words and punctuation automatically. Use fn + space for hands-free dictation.
No. Dictation is included and works system-wide, but the heart of mrmr is Agent Mode: using your voice to get real work done across apps and your Mac.
The private beta currently supports Slack, Linear, Google Calendar, Google Tasks, Google Meet, Zoom, Notion, Gmail, Cal.com, Calendly, Attio, Apple Reminders, and GitHub, plus built-in web search and on-Mac tools like file search and your browser history and bookmarks. More are on the way.
It's a real conversation. The agent asks clarifying questions, chains multiple steps across apps, reads results back, and lets you correct course mid-task, and it confirms before anything is written.
Yes. Hand bigger tasks (research, summaries, multi-step lookups) to a background sub-agent and keep talking while it works. Ask for the status anytime, and every run is saved in the app so you can come back to the result later.
The microphone only opens when you trigger mrmr; there's no always-on listening. Audio is processed in real time to power transcription and the agent; we never store recordings or use them for training.
No. Both modes need an internet connection for transcription and the agent.
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, and nothing runs in the background without your knowledge.
The beta is best suited for Mac users who work in Slack, Linear, and Google Calendar and are comfortable giving feedback on an early product.
mrmr is in private beta. Join the waitlist and we'll send access as spots open up. For the fastest path, book a 20-minute setup call.
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.
mrmr is in private beta. Pricing will be announced before paid plans launch, with founder pricing for early beta users.