Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s

Strata is free, open-source software that enables users to run Qwen 3.8 Flash Next, a 125-billion-parameter AI model, on consumer gaming PCs with graphics cards (12GB+ VRAM) at speeds of 44-94 tokens per second depending on hardware configuration. The model performs tasks like chatting, code writing, and image analysis entirely locally without data leaving the user's PC, requiring Windows or Linux and approximately 80GB of disk space. Installation is automated and can be initiated by AI coding assistants or manually through a simple setup process.