AlwaysReddy is a hotkey-activated LLM voice assistant designed for seamless, background interaction with AI models. It targets users who want to integrate AI into their workflow without context switching, offering features like voice chat, clipboard integration, and custom action extensibility.

How It Works

The assistant operates via configurable hotkeys, enabling users to initiate voice commands, interact with clipboard content, and trigger custom scripts. It supports a variety of local and cloud-based LLM and TTS engines, prioritizing a "frictionless" user experience. The architecture is designed to be extensible, allowing users to define custom actions and integrate new AI models or tools.

Quick Start & Requirements

• Install: Clone the repository and run python setup.py (Windows) or python3 setup.py (macOS/Linux).
• Prerequisites: Python, Git. GPU acceleration requires CUDA Toolkit and PyTorch with CUDA support.
• Configuration: Edit config.py and .env files for API keys and settings.
• Docs: Setup Guide, Demo Video

Highlighted Details

• Hotkey-driven interface with no GUI.
• Supports local LLMs (LM Studio, Ollama) and TTS (Piper).
• Clipboard integration for reading and writing text.
• Extensible custom action system with hotkey binding.
• Image detection for clipboard content.

Maintenance & Community

• Project creator is actively seeking feedback and contributions.
• Discord server available for community support: https://discord.gg/su44drSBzb

Licensing & Compatibility

• License not explicitly stated in the README.
• Compatibility with commercial or closed-source projects is undetermined.

Limitations & Caveats

Linux hotkey detection is limited to focused applications, hindering background operation. The codebase is described as "messy" due to ongoing development towards an extensible platform. Custom actions are experimental and subject to change.