Qwen 3.8 Flash via SSD: Max Quants on Low RAM (vs Kimi 3.1 & Qwen 4) 🤯

In this video we try to re-create the new benchmark of Qwen 4 and Kimi K3.1 with several quants of Qwen 3.8 Flash Next utilising N-Gram streaming which allows you to offload GBs of the model at no performance loss. TEST SYSTEM Inferencer v2.4.6: https://inferencer.com Q4s: https://huggingface.co/inferencerlabs/Qwen3.8-Flash-Next-MLX-Q4s Q6s: https://huggingface.co/inferencerlabs/Qwen3.8-Flash-Next-MLX-Q6s Q9s: https://huggingface.co/inferencerlabs/Qwen3.8-Flash-Next-MLX-Q9s M5 Max 128GB: https://vtudio.com/a/?a=macbook+pro M3 Ultra 512GB: https://vtudio.com/a/?a=mac+studio BUY NOW LG C2 42" Mo