Zhipu AI/Mixture of Experts

GLM-5.2

Name: GLM-5.2
Author: Zhipu AI

GLM-5.2 — Zhipu's 753B MoE flagship with 39B active and a 1M context, MIT licensed. Its predecessor GLM-5.1 shares the same expert layout.

chatcodingreasoningmultilingualmathagentictool_use

753.33B

Parameters (39B active)

1024K

Context length

Benchmarks

Quantizations

2.7M

HF downloads

Architecture

MoE

Released

2026-06-16

Layers

KV Heads

Head Dim

192

Family

glm

Quantization Options

Context length:

Quant	Bits	VRAM @ 16K	Quality
IQ2_XXS	2.38	225.6 GB 224.6 + 1.0 KV	low
IQ2_M	2.93	277.4 GB 276.4 + 1.0 KV	low
Q2_K	3.16	299.1 GB 298.1 + 1.0 KV	low
IQ3_XXS	3.25	307.6 GB 306.5 + 1.0 KV	low
IQ3_XS	3.5	331.1 GB 330.1 + 1.0 KV	low
Q3_K_S	3.64	344.3 GB 343.3 + 1.0 KV	low
IQ3_M	3.76	355.6 GB 354.6 + 1.0 KV	low
Q3_K_M	4	378.2 GB 377.2 + 1.0 KV	low
Q3_K_L	4.3	406.4 GB 405.4 + 1.0 KV	moderate
IQ4_XS	4.46	421.5 GB 420.5 + 1.0 KV	moderate
Q4_K_S	4.67	441.3 GB 440.2 + 1.0 KV	moderate
Q4_K_M	4.89	462.0 GB 461.0 + 1.0 KV	good
Q5_K_S	5.57	526.0 GB 525.0 + 1.0 KV	good
Q5_K_M	5.7	538.3 GB 537.2 + 1.0 KV	good
Q6_K	6.56	619.2 GB 618.2 + 1.0 KV	excellent
Q8_0	8.5	801.9 GB 800.9 + 1.0 KV	lossless
FP16	16	1508.2 GB 1507.1 + 1.0 KV	lossless

Select your GPU above to see speed estimates and compatibility for each quantization.

Too big for a single GPU — plan a multi-GPU deployment

Even the lightest quant needs ~226 GB. Size GPUs, replicas, TCO and scaling for a production setup. Open in Enterprise →

▸ READY TO RUN THIS?RENT BY THE HOUR

RENT A GPU AND RUN GLM-5.2 NOW

Rent on Vast.ai →Or RunPod →

Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.

Community Ratings

Loading ratings...

Benchmarks (9)

τ²-Bench99.1

GPQA Diamond89.5

AA Long Context76.7

IFBench73.3

AA Coding68.8

AA Intelligence52.6

Terminal-Bench50.8

SciCode50.5

HLE41.1

Run this model

▸Easiest way to get started·Beginners

DOCS ↗

curl -fsSL https://ollama.com/install.sh | sh

$ollama run glm:753b-q4_K_M

Tag may need adjustment — check ollama.com/library/glm for available tags.

▸ SETUP GUIDE

Auto-setup with fitmyllm CLI

Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.

pip install fitmyllmthen run fitmyllmLearn more

Auto-detect GPULive tok/s in chatSpeed benchmarks9 inference engines

HuggingFace GGUF Downloads Build Hardware

Find the best GPU for GLM-5.2

Build Hardware for GLM-5.2

QUANT	BPW	VRAM	QUALITY
IQ2_XXS	2.38	224.6 GB	65%
IQ2_M	2.93	276.4 GB	75%
Q2_K	3.16	298.1 GB	78%
IQ3_XXS	3.25	306.5 GB	82%
IQ3_XS	3.5	330.1 GB	84%
Q3_K_S	3.64	343.3 GB	85%
IQ3_M	3.76	354.6 GB	86%
Q3_K_M	4	377.2 GB	88%
Q3_K_L	4.3	405.4 GB	90%
IQ4_XS	4.46	420.5 GB	92%
Q4_K_S	4.67	440.2 GB	93%
Q4_K_M	4.89	461.0 GB	94%
Q5_K_S	5.57	525.0 GB	96%
Q5_K_M	5.7	537.2 GB	96%
Q6_K	6.56	618.2 GB	97%
Q8_0	8.5	800.9 GB	100%
FP16	16	1507.1 GB	100%