jarvis-os 0.1.22

Voice assistant with AI, voice input, and system control

Jarvis OS

Jarvis OS is a voice assistant designed to streamline and automate anything from your daily tasks to hardcore developing and everything in between. It integrates offline speech recognition, AI-powered natural language responses, and system-level command execution. Additional customization can be done by using LM studio alongside Jarvis on your own machine. Jarvis can also be used as a voice-powered CLI, accesing, organizing, and manipulating files as well as running simple bash commands.

Features

• Offline speech recognition with Vosk
• Natural voice synthesis using Yapper TTS
• Streaming from Local or Online LLM models
  • Default: Gemini 1.5 Flash
  • Local Default: Llama 3.2 1B parameters [Run LM Studio server and change URL for LLM calls in main.py to customize or change model]
• Implemented RAG system for queries that may require or benefit from web searches
• Terminal-based interface with dynamic boot and loading screens
• Executes system commands: open applications, perform searches, file management, shut down, and more

Example commands:

"Jarvis open vscode"

"Jarvis google how to bake sourdough"

"Jarvis shutdown"

"Jarvis explain quantum entanglement"

"Jarvis delete the read me file"

Reccomended before Install

It is best to run inside a python virtual enviornment:

python -m venv jarvis-env

Windows:

jarvis-env\Scripts\activate

POSIX (Max/Linux):

source jarvis-env/bin/activate

-- )

Installation

pip install jarvis-os

Configuration

Python version 3.8 or higher is required

Ensure microphone input is enabled on your system

To use an AI backend (like LM Studio), modify the API endpoint in main.py

Additional Notes

This package was initially built for Windows software. Support for POSIX is in progress (so some features work and others may not).

License

This project is licensed under the Apache License 2.0.

Author

Arvin Adeli