Is GLM-5.2 or DeepSeek-V4-Flash-0731 smaller?

Both publish quantizations; compare the table for the exact sizes.

Which uses less memory at long context?

DeepSeek-V4-Flash-0731 — its KV cache at 32K context is 43.5x smaller, because of how its attention layers cache.

Model comparison

GLM-5.2 vs DeepSeek-V4-Flash-0731

These two publish different quantization sets; the table below has the exact sizes. At long context the gap widens: DeepSeek-V4-Flash-0731's KV cache at 32K is 43.5× smaller, which usually matters more than the difference in weights.

From the file· summed bytes, KV per layer

Side by side

	GLM-5.2	DeepSeek-V4-Flash-0731
Parameters	753B	304B
Architecture	glm-dsa	deepseek4
Layers	78	43
Native context	1,048,576	1,048,576
Mixture of experts	yes, 256 experts	yes, 256 experts
Quantizations published	20	11
Smallest quantization	169.33 GiB	76.87 GiB
Q4_K_M	—	—
Licence	mit	mit

KV cache by context

the term that decides long-context viability

Context	GLM-5.2	DeepSeek-V4-Flash-0731	Ratio
4,096	0.34 GiB	0.06 GiB	5.44×
8,192	0.69 GiB	0.06 GiB	10.88×
16,384	1.37 GiB	0.06 GiB	21.77×
32,768	2.74 GiB	0.06 GiB	43.53×
65,536	5.48 GiB	0.06 GiB	87.07×
131,072	10.97 GiB	0.06 GiB	174.14×