Model facet · Hardware

GLM-5.3 GPU requirements

GLM-5.3 is a ~744B-parameter mixture-of-experts model with roughly 40B active parameters per token. Full-precision self-hosting needs on the order of 1.5TB of VRAM; aggressive 2-bit quantization brings that down to roughly 245GB. These are architecture-driven estimates, not a confirmed deployment guide — self-hosting depends on open weights, which were not yet released as of this page’s last verification.

High-memory GPUs on OneInfer

A single-GPU instance below is rarely enough for a model this size — expect a multi-GPU node. These are current listings by VRAM, for scoping a self-host deployment once weights are available.

GPUVRAM per cardListing
B300262 GBView
B200 SXM192 GBView
H200 SXM141 GBView
A10080 GBView
H10080 GBView

VRAM by precision

PrecisionApproximate VRAMNotes
FP16 / full precision~1.5 TBMulti-GPU datacenter deployment only.
FP8~750 GBHalves memory at typical accuracy cost for this model class.
2-bit quantized~245 GBThe realistic self-host floor; expect measurable quality loss.

Weights are not released yet

Self-hosting GLM-5.3 requires open weights, which Z.ai has not shipped as of this page’s last verification. Treat every figure above as an architecture-driven estimate until a real checkpoint and license are confirmed.

Ready to test the workflow?

Create account & add credits

Architecture

GLM-5.3 uses a mixture-of-experts design: roughly 744B total parameters with about 40B active per token, which is why its serving cost sits well below a dense model of comparable total size.

Frequently asked questions

What GPU do I need to run GLM-5.3?

Roughly 1.5TB of VRAM at full precision, or about 245GB at 2-bit quantization — realistically a multi-GPU datacenter node once open weights ship.

Is GLM-5.3 open source?

Open weights were promised for late August 2026 with an MIT license expected but unconfirmed as of this page’s last verification. At launch, GLM-5.3 was API-available only.

Where is the live GLM-5.3 model page?

The canonical model page with current OneInfer pricing, capabilities, and availability is /models/zai-org/GLM-5.3. This page is a focused facet of that entity, not a replacement for it.

How should I treat benchmark or price claims?

Check each claim’s provenance label and observed date. Vendor-reported and independently verified numbers are shown as separate evidence classes on this hub.

Put GLM-5.3 to work

Fund a controlled evaluation, start with a prepared prompt, and measure quality and cost on your own workload.