Instructions to use zai-org/GLM-4.5-Air with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use zai-org/GLM-4.5-Air with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="zai-org/GLM-4.5-Air")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("zai-org/GLM-4.5-Air")
model = AutoModelForCausalLM.from_pretrained("zai-org/GLM-4.5-Air")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use zai-org/GLM-4.5-Air with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "zai-org/GLM-4.5-Air"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zai-org/GLM-4.5-Air",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/zai-org/GLM-4.5-Air

SGLang

How to use zai-org/GLM-4.5-Air with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "zai-org/GLM-4.5-Air" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zai-org/GLM-4.5-Air",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "zai-org/GLM-4.5-Air" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "zai-org/GLM-4.5-Air",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use zai-org/GLM-4.5-Air with Docker Model Runner:
```
docker model run hf.co/zai-org/GLM-4.5-Air
```

When will GLM4.5 Flash be released

by WilliamKing9 - opened Jul 30, 2025

Discussion

WilliamKing9

Jul 30, 2025

I find that model in your website, it will be a good model to run locally if it is released.

deh40

Jul 30, 2025

Looking forward to it as well

ZHANGYUXUAN-zR

Z.ai org Jul 30, 2025

Flash is based on the Air model with inference optimizations, making the inference cost relatively low. It includes constraints on output tokens, with the main purpose of allowing community developers to experience the latest models for free.

huangcc

Jul 30, 2025

•

edited Jul 30, 2025

Lets' go !!!

sudo-0x2a

Aug 2, 2025

Flash is based on the Air model with inference optimizations, making the inference cost relatively low. It includes constraints on output tokens, with the main purpose of allowing community developers to experience the latest models for free.

I'm wondering if there is a plan of releasing a smaller model like the GLM4 9b? or a 16B MoE would be very interesting.

WilliamKing9

Aug 2, 2025

Flash is based on the Air model with inference optimizations, making the inference cost relatively low. It includes constraints on output tokens, with the main purpose of allowing community developers to experience the latest models for free.

I'm wondering if there is a plan of releasing a smaller model like the GLM4 9b? or a 16B MoE would be very interesting.

I also want these models, Qwen3 Coder works well on bigger projects but GLM4.5 make better results on GUI and three.js .

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment