nvidia/Qwen3.6-35B-A3B-NVFP4

A quantized inference checkpoint of Alibaba's Qwen3.6-35B-A3B mixture-of-experts language model, published by NVIDIA in its NVFP4 format for efficient deployment on NVIDIA GPUs. The checkpoint targets reduced VRAM usage and faster inference on Blackwell-era hardware while preserving the original 35B-total/3B-active architecture. Source: supabase.