GLM-5.2

GLM family · 753.0B params

1recorded run

29.7best tok/s

1verified

Evidence: 1 sourced run · 1 independent source · first recorded Sep 4, 2026. What these labels mean.

What the GLM-5.2 numbers mean

The fastest verified result on our board is 29.7 output tok/s on an CMP 170HX x10 at INT4-INT8-Mix with a 131072-token context. That is about 33.7 ms per generated token — 500 tokens takes roughly 16.8 seconds. Prompt processing on the same record runs at 193387.9 tok/s, so a 2,000-token prompt is read in about 0.0 seconds.

How much VRAM does GLM-5.2 need?

GLM-5.2 has 753.0B parameters. These estimates use the same formula as the VRAM calculator: weights, plus about 15% for a moderate-context KV cache and runtime overhead, plus a small engine floor.

Quantisation	Estimated VRAM	Smallest card that fits
Q4_K_M (4-bit)	487.9 GB	multi-GPU or CPU offload
Q8_0 (8-bit)	920.9 GB	multi-GPU or CPU offload
FP16 (16-bit)	1732.7 GB	multi-GPU or CPU offload

Long contexts and large batches raise the real number — read the smallest card as "possible with a reduced context". Measured throughput is on the benchmark board.

Fastest GPUs for this model (best tok/s)

CMP 170HX x1029.7

Models with a similar footprint

Model	Params	Q4_K_M weights
DeepSeek-V4.1-Flash	552B	310.5 GB
DeepSeek-V4.1-Flash-1017	552B	310.5 GB
GLM-5.3-Flash	321B	180.6 GB
DeepSeek-V4-Flash-Vision-Exp	305B	171.6 GB
keys-DeepSeekV4Flash-Vision-EXP-ablit	305B	171.6 GB
DeepSeek-V4-Flash-0731	304B	171.0 GB

Ranked by distance from GLM-5.2's parameter count. Full benchmark board.

All benchmarks Methodology

GPU	Model	Quant	Backend	Ctx	tok/s	prompt tok/s	VRAM GB	Source
CMP 170HX x10 · 64 GB	GLM-5.2	INT4-INT8-Mix	vllm 0.20.2	131072	29.7 ✓	193387.9	—	src