Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s
Summary
Strata enables running the Qwen3.8-Flash-Next model on consumer hardware, with detailed install steps, hardware requirements, and multi-GPU support. It includes speed benchmarks on RTX GPUs and guidance for selecting model sizes based on RAM, aiming to bring large AI inference to local machines.