CanItRun Logocanitrun.

GLM-5 744B

GLM-5 744B needs roughly 428.4GB VRAM at Q4_K_M quantization (1678.3GB at FP16). 2 GPUs we track can run it fully in VRAM at 8k context.

Z.ai744B params40B active (MoE)198k contextMITCommercial use ok

VRAM at each quantization

Assumes 8k context. KV cache grows linearly with context length.

QuantWeightsKV cacheTotal
FP161488.0 GB10.47 GB1678.3 GB
Q8744.0 GB10.47 GB845.0 GB
Q6_K558.0 GB10.47 GB636.7 GB
Q5_K_M465.0 GB10.47 GB532.5 GB
Q4_K_M372.0 GB10.47 GB428.4 GB
Q3_K_M297.6 GB10.47 GB345.0 GB
Q2_K223.2 GB10.47 GB261.7 GB

Benchmarks

GPUs that run GLM-5 744B natively (2)

Notes

Uses DeepSeek Sparse Attention (DSA) for efficient long-context. 256 routed experts, 8 active.

Hugging Face ↗Released 2026-02-17