GLM-5.3-Flash-GGUF

Overview

GLM-5.3-Flash (320B total, 18B active, multimodal) at Unsloth's Dynamic 4-bit - frontier-class chat, coding and vision in a 200 GB llama.cpp download that fits 256 GB of VRAM or unified memory. Needs a recent llama.cpp build with glm5-next support.

Good for

• agentic
• chat
• coding

Recommended hardware

≈ 256 GB VRAM to run this 321B model at the UD-Q4_K_XL quantization.

Needs 256 GB+ of VRAM (e.g. 2x 192 GB Blackwell, or 4x 80 GB H100)

Links

• View on HuggingFace
• Run on Infersec