Qwen 3.8

“Right here, we are rocking the Qwen3.8 27B 4-bit model on both machines.”

“So, I swapped GPT-5.6 with Qwen 3.8, and gave it the same kind of jobs I normally use OpenAI's models for.”

“I can use this entirely local on the mini PC for some basic task, you know, like Qwen 3.8 in like a Q4 quant.”