HAL 9000 is a local, multimodal AI agent designed to perceive via webcam, listen via microphone, process with an LLM, speak with voice cloning, and execute actions on your macOS, Windows, or Linux computer.

Can I use HAL 9000 without API keys?

Absolutely. HAL 9000 offers a 'Free Mode' that runs fully locally with Ollama for LLMs and faster-whisper for STT, requiring zero API keys and incurring no costs.

What operating systems does HAL 9000 support?

HAL 9000 boasts cross-platform compatibility, running seamlessly on macOS, Windows, and Linux with OS-aware functionality to adapt to each system's commands and features.

Does HAL 9000 remember information and past interactions?

Yes, HAL 9000 includes a Persistent Knowledge Base. It stores typed memories, auto-summarizes sessions, and allows you to upload various documents (PDFs, code, images) for retrieval and deep-reading.

HAL 9000: Local Multimodal AI Agent for Desktop Automation

HAL 9000 is a powerful, local multimodal AI agent designed to interact with your machine through vision, hearing, thought, and speech. It sees via your webcam, hears your voice, processes information using customizable LLMs (including local models like Ollama), speaks with a chosen voice, and acts on your machine using a comprehensive suite of 43 cross-platform tools. Beyond individual interaction, HAL serves as a co-work hub, facilitating collaboration with other AI agents like Claude Code, managing background tasks, orchestrating multi-agent workflows, and centralizing project artifacts and a persistent knowledge base, all controllable from a browser-based panel.

主な機能

01Multimodal interaction: Sees via webcam, hears via microphone, speaks with voice cloning, acts with 43 tools.

022 GitHub stars

03Cross-platform compatibility: Seamlessly runs on macOS, Windows, and Linux with OS-aware functionality.

04Free Mode: Fully operational with local LLMs (Ollama) and STT (faster-whisper), requiring zero API keys.

05Co-work Hub: Integrates with Claude Code for bidirectional AI collaboration, background task management, and multi-agent orchestration.

06Persistent Knowledge Base: Stores typed memories, session summaries, and allows upload of various documents for retrieval and deep-reading.

ユースケース

01Automating complex tasks and workflows on your local machine using natural language commands.

02Collaborating with Claude Code and other AI agents on software development and creative projects.

03Developing and prototyping AI solutions without relying on external API keys.