The AI companion that lives in your menu bar.
Voice-first, screen-aware, always one keystroke away. Speak to your computer and watch Odin do the work — across sessions, across tools.
Three modes, one assistant
Different work needs different presence.
Odin runs in three modes — each tuned for a different kind of task and a different amount of your attention.
Finds the thing you couldn’t.
Ask Odin a question. It sees your screen, navigates through whatever depth of UI it takes, and points you to the exact thing — out loud.
- Push-to-talk from anywhere on macOS or Windows
- Cursor lands on the precise element, not a vague area
- Spoken reply includes the path it took
Same memory. Same skills. Pick the mode by what you need to happen next.
Voice-first
Hands-on-keyboard is a habit, not a requirement. Talk, point, move on.
Screen-aware
Odin sees what you see. Answers reference the actual pixels, not guesses.
Memory across sessions
Preferences, project paths, recurring workflows — durable, not ephemeral.
Common questions
Before you sign up.
What platforms does Odin run on?
macOS and Windows at launch. It lives in your menu bar (macOS) or system tray (Windows) — no dock icon, no chrome.
How much will it cost?
Pricing isn’t locked yet. Waitlist members will get first access and early-supporter pricing before public launch.
How does Odin handle privacy?
Voice, screen captures, and memory live on your machine by default. Network calls go only to the model you’ve configured. You can audit and clear everything from the menu.
When does it launch?
Private beta is underway. Public availability follows once invites stabilize — we’ll email the waitlist before opening anything up. The roadmap tracks what’s shipping next.
Can I bring my own model / API key?
Yes. Odin is provider-agnostic — bring an OpenAI, Anthropic, or OpenRouter key and pick your model. No vendor lock-in.
Get on the list.
We email the list before public launch. No drips, no nags.