IsValorum/Qwen3.6-35B-A3B-MTP-APEX-I-NanoPlus-GGUF
Image-Text-to-Text • 36B • Updated • 201 • 2
Architectural quantization line targeting ~13 GB with solid Q4 quality and fast RAM streaming on 16GB VRAM.
Note Qwen3.6-35B-A3B calibrated to 13.03 GB (WikiText-2 PPL: 5.5244).