GLM-6 355B-A32B — VRAM, Speed & Local Setup

Name: LLM Configurator — GPU VRAM Checker
Author: LLM Configurator

Written by Jakub Rusinowski · Last updated June 26, 2026

PREVIEW — unverified. Frontier GLM-6 MoE (~355B total, ~32B active), MIT, 200K context, agentic/SWE focus. Datacenter-class to self-host (~200 GB at Q4). Estimates provisional — confirm on the Hugging Face model card.

GLM-6 355B-A32B needs about 215 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.

Specifications

Parameters	355 Billion (32B active)
Context window	200,000
Architecture	Mixture-of-Experts
Provider	Zhipu AI (Z.ai)
Licence	MIT
Specified at	Q4_K_M
System RAM	512 GB
Record updated	2026-06-26

Licence

MIT — commercial use permitted. Commercial use permitted. No usage restrictions beyond attribution.

VRAM and Speed by Quantization

Modelled on a reference NVIDIA RTX 4090 (24 GB), with no KV cache (this record has no published architecture). Speed figures are ESTIMATES from the memory-bandwidth roofline described on the methodology page, not benchmarks we ran — rows marked measured come from published or reader-submitted runs. VRAM here includes the KV cache, so it reads higher than the headline figure above, which does not.

Quant	Weights	VRAM needed	Est. speed	Fit on 24 GB
Q2_K	116.7 GB	117.5 GB	—	Won't fit
Q3_K_M	151.3 GB	152.1 GB	—	Won't fit
Q4_K_M	214.3 GB	215.1 GB	—	Won't fit
Q5_K_M	251.6 GB	252.4 GB	—	Won't fit
Q6_K	291.1 GB	291.9 GB	—	Won't fit
Q8_0	377.2 GB	378.0 GB	—	Won't fit
F16	710.0 GB	710.8 GB	—	Won't fit

Want the memory numbers alone, at every quantization level and your own context length? Use the GLM-6 355B-A32B VRAM calculator.

Buy This HardwareApple Mac Studio M3 Ultra — 512 GB VRAM · 60 W board power Deploy in the Cloud NowRunPod

or compare on Vast.ai

As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.

Recommended GPU

The cheapest catalogued GPU that runs GLM-6 355B-A32B is the Apple M3 Ultra (512 GB).

Affiliate disclosure: Some links on this page are affiliate links — if you buy through them, LLM Configurator may earn a commission at no extra cost to you. As an Amazon Associate, LLM Configurator earns from qualifying purchases.

Apple Mac Studio M3 Ultra

512 GB VRAM · 60 W board power

2026 prices are volatile — check the current listing.

Check price on Amazon

How to Run GLM-6 355B-A32B

Install Ollama, then run:

ollama run glm-6

Weights on Hugging Face: zai-org/GLM-6.

Best for: software engineering, agentic, long context, enterprise.

Can I Run GLM-6 355B-A32B on My GPU?

Other GLM-6 Sizes

GLM-6 9B — about 6 GB VRAM at Q4_K_M

GLM-6 355B-A32B — Frequently Asked Questions

How much VRAM does GLM-6 355B-A32B need?

About 215 GB at Q4_K_M — quantized weights plus framework overhead, before any KV cache. The cache grows with context length and is added on top; the table above folds it in. Apple Silicon counts unified memory toward the same figure.

Does GLM-6 355B-A32B run on an RTX 4090 (24 GB)?

No. GLM-6 355B-A32B needs about 215 GB at Q4_K_M, more than a single RTX 4090's 24 GB. It needs a larger card, several GPUs, or Apple Silicon with enough unified memory — or it runs with part of the weights offloaded to system RAM, which is much slower.

How do I run GLM-6 355B-A32B locally?

Install Ollama and run `ollama run glm-6`. That pulls the weights and starts a local OpenAI-compatible endpoint; after the download nothing leaves the machine.

What other sizes does GLM-6 come in?

GLM-6 9B (6 GB), GLM-6 355B-A32B (215 GB). Every size shares the family's training and licence; the larger ones score higher and need proportionally more memory.

← All GLM-6 models | VRAM calculator | Check your own hardware