Instructions to use axiomofmind/GLM-5.3-Flash-W4A16-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use axiomofmind/GLM-5.3-Flash-W4A16-NVFP4 with Transformers:

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

pipe = pipeline("text-generation", model="axiomofmind/GLM-5.3-Flash-W4A16-NVFP4")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("axiomofmind/GLM-5.3-Flash-W4A16-NVFP4")
model = AutoModelForMultimodalLM.from_pretrained("axiomofmind/GLM-5.3-Flash-W4A16-NVFP4", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use axiomofmind/GLM-5.3-Flash-W4A16-NVFP4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "axiomofmind/GLM-5.3-Flash-W4A16-NVFP4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "axiomofmind/GLM-5.3-Flash-W4A16-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/axiomofmind/GLM-5.3-Flash-W4A16-NVFP4

SGLang

How to use axiomofmind/GLM-5.3-Flash-W4A16-NVFP4 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 "axiomofmind/GLM-5.3-Flash-W4A16-NVFP4" \
    --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": "axiomofmind/GLM-5.3-Flash-W4A16-NVFP4",
		"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 "axiomofmind/GLM-5.3-Flash-W4A16-NVFP4" \
        --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": "axiomofmind/GLM-5.3-Flash-W4A16-NVFP4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use axiomofmind/GLM-5.3-Flash-W4A16-NVFP4 with Docker Model Runner:
```
docker model run hf.co/axiomofmind/GLM-5.3-Flash-W4A16-NVFP4
```

GLM-5.3-Flash W4A16 NVFP4

An NVIDIA ModelOpt W4A16 NVFP4 checkpoint of zai-org/GLM-5.3-Flash-BF16.

The main-model routed experts use NVFP4 weights with group size 16 and BF16 activations. Attention, shared experts, routers, embeddings, output head, and MTP weights retain their source precision.

This repository contains the Hugging Face checkpoint, not GGUF. A runtime with support for this ModelOpt W4A16 NVFP4 architecture is required.

The original model's MIT license is included. Architecture, usage, chat format, and limitations are documented in the official model card.