
omi
BasedHardware/omiAI assistant capturing audio and screen to build personal memory
PersonalAI & LLM toolingNotes & docsWebDesktopMobilePermissiveAt a glance
- stars on GitHub
- 14k
- contributors
- 367+
- last commit
- latest release, 5 hours ago
- v0.12.555+12555-macos
- license
- MIT
About
Omi is an open-source AI memory platform that connects to desktops, phones, and wearable devices. It transcribes conversations in real time, captures screen activity, creates automated summaries, and lets users query past context through an AI chat interface. It can be run via native desktop apps on macOS and Windows, mobile apps on iOS and Android, or through a web interface.
From the repository: “AI that sees your screen, listens to your conversations and tells you what to do”
Downloads
All releases on GitHub ↗Latest release: Omi Desktop v0.12.555 (candidate), . Its files were downloaded 8 times.
All files (7)
macOS
- omi-beta.dmg275 MB · 0 downloads
- omi.dmg275 MB · 0 downloads
Other files
- Omi.zip312 MB · 0 downloads
- Omi.Beta.zip313 MB · 6 downloads
- Omi.dSYM.zip111 MB · 0 downloads
- desktop-smoke-result.json1.7 KB · 1 download
- desktop-smoke-result-beta.json2.3 KB · 1 download
These files come from the project's GitHub release. DigGitHub does not check them.
Alternatives
All alternatives →| Project | Stars | Last commit | License | Worth |
|---|---|---|---|---|
| 14k | MIT | 84 | ||
| 78k (best) | AGPL-3.0 | 82 | ||
| 30k | AGPL-3.0 | 84 | ||
| 10k | MIT | 77 | ||
| 41k | MIT | 90 (best) |
▲ marks the most stars and the best worth score.
