Skip to main content
Omi captures conversations through a wearable, your phone, or your Mac. It transcribes them as they happen, then extracts memories, action items, and summaries that you can search and chat with. The apps, backend, firmware, and hardware designs are all open source.

Open source

Apps, backend, firmware, and hardware files, all in the repo

Live transcription

Speech-to-text as you speak, with speaker identification

Conversation memory

Facts and action items extracted from what you say

A platform

Apps, webhooks, chat tools, and an app store for what others build

​
The devices

Omi

Omi consumer wearable
The consumer wearable (CV1), built for everyday use

omiGlass

omiGlass
Open-source smart glasses you can build yourself
The Omi app also works without a wearable: record from your phone, or use the Mac app to capture meetings and screen context. Older developer kits are documented under Legacy & community.

​
Using Omi day to day

These docs are for builders. For end-user guides (device setup, charging, pairing, subscriptions), see the help center.

​
Where to go next

Choose your path

Pick the entry point that matches what you want to build

How Omi works

Architecture, capture, the data model, and chat

Developer API

REST API for memories, conversations, and action items

Contribute

Dev setup, PR quality bar, and community