Voice Commander: Control Your Mac with Hey Jarvis

Voice Commander is a software solution that enables users to control their Mac computers using voice commands. Unlike cloud-based AI, Voice Commander operates locally on the user's computer, eliminating the need for costly token usage, API calls, or sharing data with third-party servers. To set up Voice Commander, users simply need to install Ollama, a locally hosted AI model, and run the command "ollama pull llama3.2" to download the GPT-4 level model.

Once installed, users can begin interacting with their Mac using voice commands in a matter of seconds, without the need for an API key.

The Voice Commander software provides access to various AI models, each tailored for specific tasks. The llama3.2 model, with a 2GB size, is ideal for general use and coding. The codellama model, with a 4GB size, is perfect for generating code. The mistral 4.4GB model excels in fast reasoning, while the phi3 model, with a 2.2GB size, offers lightweight tasks. The gemma2 model, with a 5.4GB size, is designed for complex reasoning.

The potential applications for Voice Commander are vast. Users can create Retrieval Augmented Generation (RAG) systems for their documents, build voice assistants that work offline, act as code reviewers for every git commit, generate content at zero cost, and create personalized AI systems that remember user preferences.

The author of this software, Amrendra Mishra, has built over 45 tools using this stack, all of which are free, local, and open source. You can find the code and more information on his GitHub repository, github.com/amrendramishra/ai-tools, and his personal website, amrendranmishra.dev.