JARVIS AI v7.3

[EpicGlazeFire]

JARVIS AI v7.3 Release

JARVIS AI v7.3 is a major rebuild of the older JARVIS project. This version is cleaner, more modular, and designed to be easier to share, install, and expand.

What changed from the old version

The old version was mostly a personal voice assistant script with commands added directly into the main files. JARVIS AI v3 has been rebuilt into a full project structure with separate systems for voice, AI, memory, reminders, HUD, apps, search, and configuration.

Major upgrades

• New modular folder structure
• First-time setup system
• User-specific config files
• Wake-word command gate
• Cleaner voice recognition
• Edge TTS speech system
• Speech queue with priority support
• “Jarvis stop” interrupt support
• Local Ollama AI support
• Streaming AI responses
• Animated Pygame HUD
• System control commands
• App launching
• Weather support
• Persistent memory
• Memory-based recommendations
• Persistent reminders
• Web search summaries
• Sports result and prediction helper
• Share-ready defaults
• Installer support through Inno Setup

Better privacy and sharing

This version separates personal data from the main source code. User settings, memory, and reminders are stored locally in the data folder, so the bot can be shared without including personal preferences, reminders, or location data.

Installer version

This release includes an installer build for easier setup on Windows. Users should still install Ollama and pull the configured AI model before using AI features.

Recommended model:

ollama pull phi3

<hr /><em>This discussion was created from the release <a href='https://github.com/EpicGlazeFire/JARVIS-OS/releases/tag/v7.3'>JARVIS AI v7.3</a>.</em>

[ptunc4834-hue]

docker pull ghcr.io/tramakrishna3012/jarvis-frontend:main

[EpicGlazeFire]

That Docker command is for pulling some container image from GitHub Container Registry. My JARVIS build isn’t distributed through Docker right now — it’s a Windows app built with PyInstaller/Inno Setup. Use the setup .exe from my GitHub release instead.