Qwen 3.8 Flash on 2x DGX Spark right now.
I'm blown away by this one. It feels like the model I've been chasing since the days when we experienced Openclaw, when Opus was actually good. Mac Studio M5 Ultra 256GB Crowd is asking the practical question: will the unified memory
Qwen 3.8 27B is already agentic lore.
So we ran two Qwen 3.8 27B models in one local build. We stacked the Qwen models using the new Windows Hermes Agent desktop bot mode running on our 128 GB Ryzen AI Max+ processor based agent computer. See how Ryzen AI Max+
Model: qwen-3.8-flash-next
Prompt: make me a game that plays like geometry dash Compute: 2x DGX Sparks Alibaba_Qwen NVIDIAAI
Build a beach simulator using Qwen 3.8 and GLM-5.3 Flash
Qwen 3.8 used about 2M tokens and was free because I ran it locally GLM-5.3 Flash used about 2-3M tokens and was free on OpenCode
You have reached the end of the archive
All of qwen38