Running a 125B MoE on a Single RTX 4090: Qwen3.8-Flash-Next Hybrid Inference Report
A community report claims Qwen3.8-Flash-Next, a 125B MoE model, runs on a single RTX 4090 with system-RAM offload. Here are the verified conditions and caveats.
#hybrid-inference
TAU HOME stories tagged hybrid-inference.