Hermes Glasses

What Hermes Glasses does

Hermes Glasses is an iPhone app for talking to an AI through smart glasses, either Meta Ray-Ban or AiSee and other Realtek-based glasses (RTL8773D plus RTL8735B reference design). Speech is transcribed live with Apple's on-device recognition, so no audio leaves the phone for speech-to-text. Finished utterances go to an AI provider you configure with your own API key (Direct mode) or to a Hermes Agent running on a Mac (bridge mode), and the answer is spoken back.

Beyond voice, it captures photos from the glasses camera for questions like what am I looking at, runs on-device YOLO object detection in its Object Snap feature, starts walking or driving navigation without an AI round trip, and keeps a private on-device record of people you meet. On Ray-Ban Display glasses, transcription and replies appear on the lens, while Realtek glasses use a temple button mapped to actions. Without glasses, a phone mode runs the features from the iPhone alone with a simulated lens. The README calls it a standalone, MIT-licensed project, and the Hermes bridge is optional.

Key features

• Live transcription using Apple's on-device speech recognition
• Direct mode with your own API key, or bridge mode to a Hermes Agent running on a Mac
• Glasses camera vision queries such as what am I looking at
• Object Snap with on-device YOLO detection and an Object Log
• Voice-started walking or driving navigation with a turn-by-turn banner on the lens
• Phone mode with a simulated lens for use without glasses

When to use it

• Asking a Hermes agent questions hands-free while wearing smart glasses
• Saving a photo and a spoken note about someone you just met
• Photographing work every few seconds and checking it against an imported procedure

Who it is for: iPhone users with smart glasses, or who want a phone-only mode, and who want a voice and vision assistant backed by an AI provider or Hermes Agent.

How it fits with Hermes Agent

It supports Hermes Agent as an optional bridge mode for agentic, tool-using conversations, alongside a direct provider mode that uses your own API key.

Requirements: An iPhone; either an API key for an AI provider (Direct mode) or a Hermes Agent running on a Mac (bridge mode); smart glasses are optional because phone mode works without them

FAQ