How much VRAM does GLM-4.5 (MoE) need?

GLM-4.5 (MoE) requires approximately 0.0 GB VRAM at Q4_K_M quantization with 1K context and 10% overhead. At Q8_0 it needs ~0.0 GB, and at FP16 ~0.0 GB. Use the PCPARTGUIDE VRAM Calculator for exact requirements at any context length and overhead setting.

Can I run GLM-4.5 (MoE) on a consumer GPU?

GLM-4.5 (MoE) is a 355B parameter model and requires datacenter or multi-GPU setups. At Q4_K_M, it needs approximately 0.0 GB VRAM, which exceeds consumer GPU limits. It can run on H100 (80 GB) or A100 (80 GB) datacenter GPUs, or across multiple consumer GPUs with tensor parallelism.

What is GLM-4.5 (MoE)'s context window?

GLM-4.5 (MoE) supports a maximum context window of 204,800 tokens. The default context is 4,096 tokens. Larger contexts require proportionally more VRAM for the KV cache — use the VRAM Calculator to estimate exact requirements at your target context length.

What quantization level should I use for GLM-4.5 (MoE)?

For GLM-4.5 (MoE), start with Q4_K_M (~0.0 GB) if VRAM-limited. Q8_0 offers better quality at ~2× the VRAM. If you have enough VRAM, Q6_K (~1.5× Q4 VRAM) provides a good balance. For models this large, Q3_K_M may be necessary to fit on consumer hardware.

How does GLM-4.5 (MoE) compare to similar models?

GLM-4.5 (MoE) is a 355.0B MoE model from the GLM family. Browse other GLM models on the PCPARTGUIDE directory for comparable alternatives with different parameter counts and VRAM requirements. Use the Model Transparency scores to assess documentation quality and reproducibility.

GLMMoEMIT

GLM-4.5 (MoE)

355.0B

Parameters

32.0B

Active

200K

Max Context

MoE

Architecture

—

Released

Text

Modality

About GLM-4.5 (MoE)

GLM-4.5 (MoE) is a mixture-of-experts (MoE) transformer language model from the GLM family, containing 355B parameters across 64 layers. It has 355B total parameters loaded into VRAM with 32B active per token. It supports up to 205K tokens of context with a hidden dimension of 7168 and 8 KV heads for efficient grouped-query attention (GQA). MIT license. MoE. 200K context. Server class.

ResearchEnterprise

Technical Specifications

Total Parameters355.0B

Active Parameters32.0B per token

ArchitectureMixture of Experts

Total Experts32

Attention TypeGQA (MoE)

Hidden Dimensiond = 7,168

Transformer Layers64

Attention Heads56

KV Headsn_kv = 8

Head Dimensiond_head = 128

Activation FunctionSwiGLU

NormalizationRMSNorm

Position EmbeddingRoPE

System Requirements

Estimated VRAM at 10% overhead for different quantization methods and context sizes.

Quantization	1K ctx	195K ctx	200K ctx
Q4_K_M0.50 B/W ~97% of FP16	183.7Cluster / Multi-GPU	232.3Cluster / Multi-GPU	233.5Cluster / Multi-GPU
Q8_01.00 B/W ~100% of FP16	367.2Cluster / Multi-GPU	415.8Cluster / Multi-GPU	417.0Cluster / Multi-GPU
F162.00 B/W Reference	734.2Cluster / Multi-GPU	782.8Cluster / Multi-GPU	784.0Cluster / Multi-GPU

Fits 24 GB consumer GPU

Fits 80 GB datacenter GPU

Requires cluster / multi-GPU

Other GLM Models

View All

GLM-5.1 (MoE)

Params

754.0B

Layers

Context

128K

32.0B active

Find the right GPU for GLM-4.5 (MoE)

Use the interactive VRAM Calculator to see exactly how much memory you need at any quantization level, context length, and overhead setting.

Open VRAM Calculator Best GPU Guide VRAM Guide