Rabbit R1 · Apps

Use the rotating Rabbit Eye camera for vision tasks

The R1's rotating camera, nicknamed the "Rabbit Eye," lets the assistant see what you're pointing it at — useful for asking questions about objects, text, or scenes in front of you.

What you should know

  • Point the Rabbit Eye at an object, document, or scene, then press and hold PTT to ask a question about what it sees.
  • The camera rotates to switch between facing away from you and facing you, depending on the task.
  • Combine voice and camera together, for example holding PTT while asking the assistant to describe or identify something in view.
  • Vision requests are processed online, the same as voice requests, so they need an active network connection.

Why it matters

The Rabbit Eye is what differentiates the R1 from a purely voice-based assistant — it lets you ask about the physical world around you, not just questions typed or spoken in the abstract.

Best practices

  • Hold the device steady for a moment while the camera captures the scene so the response is based on a clear view.
  • Make sure there's enough light on the subject — vision features rely on the camera actually being able to see clearly.

Common problems and fixes

Camera doesn't seem to rotate to the expected position
Give the motor a moment to finish moving before speaking, and avoid physically blocking the lens while it rotates.
Vision responses seem inaccurate
Try moving closer, improving lighting, or reframing the shot so the subject fills more of the frame.

Tips and tricks

  • Ask a specific question about what's in view (e.g. "what is this?") rather than a vague request, since the camera and LAM work together on the request you actually spoke.

Frequently asked questions

Is the Rabbit Eye always recording?

No — like the microphone, the camera is tied to your Push-to-Talk request rather than continuously recording in the background.