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 Settings

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
```

i love GLM! but please consider longer context length

#10

by mahmood36 - opened Aug 9, 2025

Discussion

mahmood36

Aug 9, 2025

im merely an enthusiast, my words might mean nothing, i cant even run the model locally.
but recently i tried the model through openrouter and woah boy, it outperformed every other model including closed source ones. this model is very reminiscent of gemini 2.5 in its early days (which later got nerfed). the only thing missing is a longer context length and it would be SOTA in every way. in my use case (which is help with academic study and synthesis of math/cs), it was brilliant and outperformed all models including the recent qwen3, but the most recent qwen3 supports a context length of up to 1m, and even offers that in very small models that i can run locally.

thank you for your work! please stay open source! you are our best hope!!

mahmood36

Aug 9, 2025

i would also like to mention out that even though i have a 4090, i cant run the model locally at a reasonable quant. tried q4 but that requires 80+ gb of memory excluding kv-cache context.

this is not a complaint, merely a humble wish, so dont feel criticized because even in its current state, i wouldnt have expected an open source to be that good in the first place.

thank you again! i appreciate your work!

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