v1.60.4 · macOS · Apple Silicon

A meeting
note-taker
that talks back.

OpenOats sits next to your call, transcribes both sides in real time, and searches your own notes to surface things worth saying — right when you need them.

Audio never leaves your Mac
Open source · MIT
how it works

Three moves. No recording studio required.

01

Start a call, hit Live.

OpenOats opens quietly in a corner. The window is hidden from screen-sharing by default — so it stays between you and your notes.

02

It transcribes both sides.

Speech recognition runs entirely on your Mac using Apple's Speech framework. Not a byte of audio ever leaves the device.

03

Your notes talk back.

When the conversation hits a question, a decision point, a claim worth backing up — OpenOats searches your notes and surfaces the exact line you meant to remember.

privacy by default

Nothing important ever leaves your Mac.

Transcription runs locally, full stop. Pair it with Ollama and everything — audio, notes, suggestions — stays on your machine. Or bring your own API key. Your data, your call.

Audio stays on device
Apple's on-device Speech framework — always.
API keys in your Keychain
Never synced. There is no us, actually.
Invisible to screen shares
Hidden from capture by default. Nobody sees it.
Fully open source
MIT license. Read every line. Fork it.
no cloud
no server
your mac
offline ✓
audio
both speakers · 16khz
local transcription
Apple Speech · on-device
your notes
embedded locally
what else

Built by someone who actually hates cloud meeting tools.

Invisible to the other side

The window is hidden from screen sharing by default. They don't see it. You do.

Fully offline, if you want

Pair with Ollama and nothing touches the network. Runs 100% on your Mac.

Bring your own LLM

OpenRouter for GPT-4o, Claude, Gemini — or Ollama for Llama, Qwen, Mistral.

Live dual-speaker transcript

See both sides as it happens. Copy with one click. Plain text, no lock-in.

Your notes, searchable

Point it at a folder of .md or .txt files. It chunks, embeds, caches locally.

Auto-saved sessions

Every conversation becomes a plain-text transcript and structured log.

Sound prepared. Because you are.

Free, open source, built for macOS. Takes about a minute to set up. The first run downloads the on-device speech model — after that, you can pull the wifi cable and it still works.

macOS 15+ · Apple Silicon·built in Swift