Olares is a personal cloud operating system that runs AI agents and large language models on hardware you own.
The problem Olares addresses is the privacy and cost concerns of relying on cloud-based AI services that store sensitive personal data on third-party servers and charge by usage. It solves this by bringing AI computation to your own hardware through a Kubernetes-based platform. The approach lets you run open-source AI applications and models locally while maintaining cloud-like accessibility through a web browser interface. You can manage the system in plain language, and it handles compute pooling, storage access, networking, and application deployment.
Olares suits individuals and small teams who want to run AI agents with local language models while retaining full control over their data and hardware. It works on Linux machines, including bare metal, virtual machines, and Raspberry Pi, with dedicated installation paths for Windows and macOS. The platform includes built-in applications for file management, secure storage, application marketplace access, and system control. It provides GPU acceleration management with time-slicing and memory-slicing modes, unified access to local files and external storage like cloud services or SMB/NFS shares, and private networking with VPN and reverse proxy capabilities so applications get HTTPS endpoints without manual port exposure.
The project shows consistent development activity with regular updates to its core functionality and documentation. The codebase is actively maintained with ongoing improvements to system stability and feature expansion. Contributors engage with the community through documentation and issue resolution. The project maintains multilingual documentation and actively develops new capabilities for AI workload management and system administration.
