Model comparison

GLM-Z1-9B-0414 vs Qwen3-VL-8B-Instruct-abliterated-v1

At Q4_K_M, Qwen3-VL-8B-Instruct-abliterated-v1 is the smaller download — 5,027,785,792 bytes against 6,166,574,464.

From the file· summed bytes, KV per layer

Side by side

GLM-Z1-9B-0414Qwen3-VL-8B-Instruct-abliterated-v1
Parameters9.4B8.8B
Architectureglm4qwen3vl
Layers40
Native context32,768
Mixture of expertsnono
Quantizations published4748
Smallest quantization2.98 GiB1.97 GiB
Q4_K_M5.74 GiB4.68 GiB
Licencemitapache-2.0

KV cache by context

the term that decides long-context viability
ContextGLM-Z1-9B-0414Qwen3-VL-8B-Instruct-abliterated-v1Ratio
4,0960.16 GiB
8,1920.31 GiB
16,3840.63 GiB
32,7681.25 GiB
65,5362.50 GiB
131,0725.00 GiB