Skip to content

Releases: rayl15/OpenVision

v1.2.1 - Bug Fix

12 Feb 05:37

Choose a tag to compare

Bug Fix

  • Fixed always-listening bug: App no longer continues listening and speaking after conversation ends
  • Properly stops voice service when session ends (especially when wake word is disabled)
  • Prevents stale commands and responses from being processed after session timeout

Previous in v1.2.0

  • Wake word "ding" sound effect
  • Thinking loop sound while AI processes
  • Connection reliability improvements (auto-retry, longer timeout)

v1.2.0 - Sound Effects

11 Feb 16:49

Choose a tag to compare

What's New

Sound Effects 🔔

  • Wake word ding: Hear a satisfying "ding" when the wake word is detected
  • Thinking indicator: Soft "pop" sounds play while the AI is processing
  • Sounds play every 3 seconds during thinking to indicate activity
  • Respects the "Activation Sound" toggle in Voice Settings

Connection Improvements

  • Auto-retry (3 attempts) for more reliable initial connections
  • Increased connection timeout for slower networks
  • Uses AudioToolbox for sounds (no audio session conflicts)

Full Changelog

v1.1.0...v1.2.0

v1.1.0 - TTS Voice Selection

11 Feb 11:40

Choose a tag to compare

What's New

TTS Voice Selection

You can now choose your preferred text-to-speech voice for AI responses!

  • Settings → Voice Control → TTS Voice
  • Browse all available voices (Premium, Enhanced, Default)
  • Preview voices before selecting with the play button
  • Your choice is saved and used for all AI responses

Tip

Download premium voices in iOS Settings → Accessibility → Spoken Content → Voices for the most natural sound.


Full Changelog: v1.0.0...v1.1.0

v1.0.0 - Initial Release

10 Feb 09:02

Choose a tag to compare

OpenVision v1.0.0

The first public release of OpenVision!

Features

  • Dual AI Backend Support

    • OpenClaw: Wake word activation, 56+ tools, WebSocket connection
    • Gemini Live: Real-time voice + vision streaming
  • Smart Voice Control

    • Wake word activation ("Ok Vision")
    • Barge-in support - interrupt AI anytime
    • Conversation mode for follow-ups
  • Meta Ray-Ban Glasses Integration

    • Photo capture on voice command
    • Live video streaming (1fps)
    • Seamless registration via Meta AI app
  • Production Ready

    • Auto-reconnect with exponential backoff
    • Network monitoring
    • Secure credential storage

Getting Started

See README for setup instructions.

Requirements

  • iOS 16.0+
  • Meta Ray-Ban smart glasses (optional)
  • OpenClaw instance or Gemini API key