arrakis
EN
Sign in Start renting
Menu

Arrakis / Models / GLM-4.5-Air

GLM-4.5-Air GPU requirements

GLM-4.5-Air has 110B parameters. To serve it, plan for about 265 GB of GPU memory in 16-bit, 133 GB in 8-bit or 66 GB in 4-bit, including about 20% for the KV cache and activations.

It is a mixture of experts: 128 experts, 8 of them active for each token. All of them stay in GPU memory, so plan memory for all 110B parameters.

Deploy this model Check another model

What it takes to serve it

GPU memory to run it for inference, with the smallest GPU setup that holds it and the lowest price on Arrakis right now.

PrecisionMemory and the smallest setup
16-bit265 GB: 2 × H200, from $6.88 per hour for all 2
8-bit133 GB: one H200, from $3.44 per GPU-hour
4-bit66 GB: one H100 or A100, H100 from $2.07 per GPU-hour

What it takes to fine-tune it

QLoRA trains adapters on a 4-bit frozen model, LoRA on a 16-bit one, and full fine-tuning updates every weight. Figures are before activations, which grow with batch size and sequence length.

MethodMemory and the smallest setup
QLoRA77 GB: one H100 or A100, H100 from $2.07 per GPU-hour
LoRA265 GB: 2 × H200, from $6.88 per hour for all 2
Full fine-tuning1,768 GB: 8 × B300, from $36.80 per hour for all 8

About the model

From its card and files on Hugging Face. Memory figures here follow the GPU memory guide.

Parameters
110B
Experts
128
Experts for each token
8
Context window (tokens)
131,072
Published in
BF16
License
mit
Released
2025-07-20
Weights
zai-org/GLM-4.5-Air

Deploy it on Arrakis

Open Models in the console, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine. Or run it yourself on any machine with enough GPU memory:

vLLM
vllm serve zai-org/GLM-4.5-Air --tensor-parallel-size 2
Deploy this model How deploying works

Questions

Live prices for every GPU are on the GPU prices page.

How much GPU memory does GLM-4.5-Air need?
To serve it, about 265 GB in 16-bit, 133 GB in 8-bit, 66 GB in 4-bit, including about 20% for the KV cache and activations. To fine-tune it, about 77 GB with QLoRA, 265 GB with LoRA and 1,768 GB for full fine-tuning with Adam, before activations.
Can I run GLM-4.5-Air on one GPU?
Yes, quantized. In 8-bit it needs about 133 GB, which fits on one H200. In 16-bit it needs 2 H200 GPUs.
What does it cost to run GLM-4.5-Air?
In 16-bit it needs 2 H200 GPUs; in 4-bit, one H100 or A100. Each setup on this page shows the lowest price per GPU-hour on Arrakis right now, and you pay by the hour from a prepaid balance.
How do I deploy GLM-4.5-Air on Arrakis?
Open Models in the Arrakis console, paste zai-org/GLM-4.5-Air, choose vLLM, SGLang or llama.cpp, and rent a GPU it fits on. You get an OpenAI-compatible endpoint with its own API key, and root SSH to the machine.

More models

Every model, and a check for any other.