Olares Launches Olares One: Privacy First Personal Desktop Cloud Server and AI Workstation

Olares One enters the market with an ambitious proposition. Instead of treating AI as a service delivered from the cloud, it positions itself as a personal cloud server and AI workstation that keeps data private, while turning local files into a personal assistant, signaling confidence in a future where consumers own the infrastructure that powers their AI experiences.

A personal AI built from your digital footprint

Olares frames the current state of AI as a tradeoff: powerful tools like Perplexity and ChatGPT offer convenience, but at the cost of privacy. These services rely on access to user queries, files, and behavioral data, turning personal information into training inputs for remote models. Olares One seeks to eliminate that compromise by building a digital twin of each user based entirely on local data.

The device analyzes documents, messages, and notes stored on your PC, converting them into a searchable knowledgebase that remains on-device. From this, it trains an assistant that learns personal writing style, routines, and preferences. Instead of generic outputs, the system aims to deliver context aware responses shaped by the user’s actual work and habits, without relying on the cloud.

Privacy minded, but designed for everyday use

Privacy focused hardware often comes with usability trade-offs, but Olares is trying to avoid that perception. Olares One runs an open-source operating system that behaves like a normal PC, allowing users to browse, download, and run applications with familiar workflows. It requires no particularly specialized setup.

It also supports remote access, enabling users to connect existing Windows devices and run their software through Olares One as if installed locally. This blending of environments may lower the barrier for users who want better performance without abandoning their tools.

Security features are integrated into the core architecture. Built in sandboxing prevents applications from accessing data without permission, creating a baseline level of isolation even if downloaded apps behave unpredictably. The company’s message is that users get stronger privacy protections with minimal friction.

Hardware aimed at AI speed rather than storage

Olares One is marketed as both a privacy oriented personal cloud and a high-performance AI workstation. It uses an NVIDIA GeForce RTX 5090 Mobile GPU with 24GB of memory, paired with an Intel Core Ultra 9 275HX processor and 96GB of RAM. This configuration is designed to support local processing of large datasets and multiple AI models simultaneously.

According to Olares, benchmark scores place it ahead of other personal AI computing systems in tests involving multimodal and lightweight models including GPT, Llama, and Qwen. The system is not just about running models quickly, but running several of them at once without resorting to cloud servers.

A key architectural feature is Time Sharing Mode, which allocates GPU resources dynamically across active applications. Olares claims this improves efficiency during multitasking and allows intensive workloads to run without dramatic performance drops.

Pricing and availability

The launch of Olares One reflects a broader shift in thinking about AI infrastructure. As AI becomes more integral to work and communication, dependence on cloud platforms raises questions about privacy, cost, and long-term control. Olares is betting that consumers and professionals will increasingly value local ownership of their tools and data.

Olares One launches on Kickstarter at US $2999. The price aims to be competitive with high end PCs and specialized workstations while offering significantly more power for AI workloads.

With its emphasis on performance, privacy, and ease of use, Olares One is targeted at early adopters who want fast AI capabilities without giving up control of their data. Whether it sparks a wider shift toward personal AI hardware remains to be seen, but it tees up a debate that is only beginning.