

Nice setup. I’m running Qwen3.8 27B at 17GB with a 32k context window on a 3090. It does most of what I need for everyday use and its all local, private, no internet connection needed.
The gap between running an LLM locally and cloud API is closing fast.

They got the call.