As AI models gain popularity, startups are competing to build software layers on top of them. One of the new projects in this space is Osaurus, an open-source LLM server designed for Apple devices. It allows users to freely switch between local or cloud-based AI models, as well as keep files and tools securely on their device. This is reported by Techcrunch.com reports .
The Osaurus project was developed by engineer Terence Pae, who previously worked at Tesla and Netflix. According to him, the idea started with creating a local AI assistant capable of working with the user's personal files, browser data, and system settings. Currently, Osaurus can work flexibly with both cloud providers like OpenAI and Anthropic, as well as locally installed models.
The main advantage of this system is that users can choose the AI model that suits their needs and manage them through a single interface. Unlike other tools like OpenClaw or Hermes, Osaurus provides a user-friendly interface for non-technical users. Additionally, to ensure security, all processes are executed in a virtual sandbox environment.
It should be noted that running AI models on a local computer requires significant resources. For stable system performance, at least 64 GB of RAM is required. For running larger models like DeepSeek v4, Terence Pae recommends using systems with 128 GB of RAM.
Read “Zamin” on Telegram!
