Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
hbar.systems describes itself as an “intelligence commons,” organizing intelligent systems around areas such as research, music, code, economics, and identity. It is not a single AI app in the traditional sense, but an experimental network made up of multiple nodes: BrainFoundry serves as the protocol layer, hbar.brain represents a personal Brain, and hbar.casino showcases a public experiment where multiple AI models play Texas Hold’em in real time.
The clearest AI capability shown on the site comes from hbar.casino: Claude, GPT-4, Gemini, Grok, DeepSeek, Llama, and other models continuously play at a 6-player table, while users can watch the revealed cards, betting, and game-theory panels. This is closer to multi-agent behavior observation and game theory research than a productivity tool. Another central concept is “one brain per person”: a user’s documents, memories, and models are deployed on servers they own, then connected to other Brains through a federated protocol that does not require a central authority. BrainFoundry is described as an intelligence engine that includes agents, orchestration, and recursive learning.
The captured page does not disclose pricing, free quotas, trials, payment methods, or a formal registration process. The site is labeled V0.1 PUBLIC, suggesting that it is currently more of an early public experiment. For regular users, there is not enough information to confirm whether they can directly create a personal Brain, whether self-hosting is required, whether installation documentation exists, or what level of technical expertise is needed.
Its strength lies in a clear philosophy: it emphasizes sovereign control over personal cognitive assets, documents, memories, and models, while avoiding capture by centralized platforms. The multi-model poker experiment also has some research and viewing value. The downside is that the product boundaries are unclear, and there is no explicit disclosure of APIs, SDKs, privacy policy, stability, or support channels. It feels more like a researcher-led system prototype than an AI SaaS product with completed commercial validation.
It is best suited for developers, researchers, or advanced users interested in local AI, personal knowledge bases, federated protocols, multi-agent systems, and decentralized intelligence infrastructure. It is less suitable for teams looking for a Chinese interface, out-of-the-box customer support, and clearly defined pricing plans. The page does not state how accessible it is from China. Since it involves models such as Claude, GPT-4, Gemini, and Grok, the actual experience may depend on the user’s network and account environment. Alternatives to watch include Obsidian, Mem.ai, Personal AI, Open WebUI, LangChain, and AutoGen.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on hbar.systems official site.
hbar.systems is an Slovakia AI Apps provider. TG4G tracks its product information, an overall rating of 4.0/10, and a China-accessibility score of Workable. Click "Visit Official Site" to reach hbar.systems directly.