Osaurus brings both local and cloud AI models to your Mac

Osaurus is an open-source LLM server designed exclusively for Apple devices, allowing Mac users to seamlessly integrate local and cloud-based AI models. It addresses a growing need for flexible AI solutions that prioritize user data privacy and control. The platform evolved from the concept of a desktop AI companion, aiming to provide a personal AI experience without relying solely on cloud services.

The system enables users to choose their preferred AI models, whether hosted locally or via cloud providers like OpenAI and Anthropic, while keeping their files and tools on their own hardware. This "harness" approach offers a unified interface for various AI models, tools, and workflows, distinguishing itself from developer-centric alternatives by offering an intuitive user experience and robust security measures through hardware-isolated virtual sandboxes.

While running AI models locally can be resource-intensive, requiring substantial RAM, the Osaurus team anticipates that local AI requirements will diminish as the technology advances. The platform currently supports a diverse array of models, including MiniMax M2.5, Gemma 4, Llama, and DeepSeek V4, alongside Apple's on-device foundation models and Liquid AI's LFM family. It also integrates with major cloud AI services like Gemini and OpenAI.

Beyond model flexibility, Osaurus functions as a full Model Context Protocol (MCP) server, offering access to its tools for any MCP-compatible client and shipping with over 20 native plugins for various macOS functionalities. With over 112,000 downloads, Osaurus is gaining traction by offering a unique feature set and a user-friendly approach compared to competitors. The company is exploring expanding its services to businesses, particularly in sectors like legal and healthcare, where local LLMs can effectively address critical privacy concerns. This move could potentially reduce the demand for large AI data centers by empowering on-premise AI deployment.