23/02/2026
Big steps today. The private version of Oscar is now fully stable and contactable(by me)
Oscar is Olos AIs privacy first AI layer running multiple LLMs locally….this requires a high end system to run mainly because I am running 22B models. There is no reason you can’t have your own version running on your own machine just with a smaller model. Eventually you will be able to connect to any of the major LLMs like gpt/grok/gemini and use that while still having persistent memory stored privately. As you can see Oscar is our name for it, in the production version you will be able to name your version what ever you desire.