Comment by replwoacause

Sounds pretty beefy. What kind of local LLM is that thing capable of running? Does it open up real alternatives to cloud providers like OpenAI and Claude, or are the local models this hardware is capable of running still pretty far behind?