Real-time meeting copilot for Mac and Windows
Say the right thing, live.
Repeeter listens to the other side of your call, notices questions and doubts, and shows you what to say in a floating window that screen sharing never sees. It runs on your computer, with your own AI keys or fully local models.
macOS 13 or later on Apple Silicon · Windows 10 or later. Free with your own keys.
Let me make sure I bring the right people in.
Usually that is whoever signs, whoever uses it every day, and whoever owns the budget line.
Security review tends to gate procurement, so I would loop them in early rather than at the end.
Who would you want on that call from your side?
What it does
Built for the second after they ask.
Hears the meeting, not just you
System audio is their side, your microphone is yours. Two channels, so it knows who said what and only ever answers them.
Answers from your notes
Drop your product sheet, pricing, objections or your résumé into a folder. Every answer is grounded in it, and anything missing gets a Check: line instead of a guess.
Invisible when you share
The overlay is excluded from screen capture on macOS and Windows. Zoom, Meet and Teams see your slides. You see the answer.
Written to be said
Spoken English that mirrors how you actually talk, not bullet points to read off. Code, when asked, arrives in a copyable block.
Anything on screen
A question in the meeting chat, a slide, an error, a snippet. One shortcut captures the screen and answers it.
Open to your other tools
A local MCP server lets Claude Desktop, Cursor or ChatGPT read the transcript and push their own suggestions into the overlay.
How it works
Four steps. Then it stays out of your way.
- 1
Install
Mac: drag Repeeter to Applications. Windows: run the installer. One short setup: choose cloud keys or local models, allow the microphone, test it.
- 2
Add your notes
Any Markdown or text in the context folder is loaded into every answer, and reloaded the moment you save.
- 3
Press Start
Join the call in any app. Their words appear on the live line. Questions get answered a second or two after they are asked.
- 4
Say it
Read it or paraphrase it. ⌘↵ answers the last thing they said, ⌘⇧C asks Repeeter anything, End saves the session to your disk.
Your AI, your choice
Cloud keys, local models, or a mix.
Cloud with your keys
Fastest and most accurate
OpenAI for live transcription. Claude, GPT, or any model on OpenRouter for answers. You pay the providers directly, usually a few cents per meeting. Keys are stored encrypted on your Mac and never touch our servers.
Local on this Mac
Nothing leaves the machine
Whisper for transcription on your GPU and a model through Ollama for answers. No per-minute cost, no key, and the overlay tells you so. Pick the model that fits your machine's memory.
Technical calls
Code when the question needs code.
Switch the meeting type to Technical and answers come engineer to engineer: the trade-off first, then a minimal, runnable snippet in a block with its own copy button, then the edge case worth mentioning. A screenshot of code or an error in the chat gets the same treatment.
Make the consumer idempotent: key each event by its delivery id and reject repeats at the database.
CREATE TABLE processed_events ( event_id text PRIMARY KEY, seen_at timestamptz NOT NULL DEFAULT now() );INSERT INTO processed_events (event_id) VALUES ($1)ON CONFLICT (event_id) DO NOTHING RETURNING event_id;
If the insert returns a row, process it; if nothing comes back, it is a duplicate.
Private by design
The transcript lives on your disk.
Audio goes to the transcription provider you chose, or nowhere at all with local Whisper. Transcript text goes to the answer model you chose. Sessions are saved as plain files on your computer. Repeeter's servers hold your account and nothing from your meetings.
- Keys encrypted by the operating system, in your own settings file
- No audio or transcript ever stored by us
- Local mode: no network at all during a call
- You decide what to tell participants; we remind you to
Your next call is a good time to try it.
Free with your own keys or local models. Setup takes a few minutes, and the first answer tells you whether it is for you.