Instructions to use cerebras/GLM-4.6-REAP-268B-A32B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cerebras/GLM-4.6-REAP-268B-A32B with Transformers:

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

pipe = pipeline("text-generation", model="cerebras/GLM-4.6-REAP-268B-A32B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("cerebras/GLM-4.6-REAP-268B-A32B")
model = AutoModelForCausalLM.from_pretrained("cerebras/GLM-4.6-REAP-268B-A32B")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use cerebras/GLM-4.6-REAP-268B-A32B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cerebras/GLM-4.6-REAP-268B-A32B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cerebras/GLM-4.6-REAP-268B-A32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/cerebras/GLM-4.6-REAP-268B-A32B

SGLang

How to use cerebras/GLM-4.6-REAP-268B-A32B 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 "cerebras/GLM-4.6-REAP-268B-A32B" \
    --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": "cerebras/GLM-4.6-REAP-268B-A32B",
		"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 "cerebras/GLM-4.6-REAP-268B-A32B" \
        --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": "cerebras/GLM-4.6-REAP-268B-A32B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use cerebras/GLM-4.6-REAP-268B-A32B with Docker Model Runner:
```
docker model run hf.co/cerebras/GLM-4.6-REAP-268B-A32B
```

Help us Cerebras - you're our only hope

by AImhotep - opened Dec 29, 2025

Discussion

AImhotep

Dec 29, 2025

Your GLM-4.6 25% REAP version is out of this world! Seriously - very good job, thank you!

Can I ask for support of GLM-4.7? I've tried to add new model to your reap repo on GitHub but unfortunately there is something that I've missed and cannot do reap by my own.
Thank you!

lazarevich

Cerebras org Jan 12

•

edited Jan 12

@alecm-cerebras we've just uploaded 4.7 REAPs:

25% FP8: https://hf.co/cerebras/GLM-4.7-REAP-268B-A32B-FP8
25% BF16: TBD

40% FP8: https://hf.co/cerebras/GLM-4.7-REAP-218B-A32B-FP8
40% BF16: https://hf.co/cerebras/GLM-4.7-REAP-218B-A32B

Enjoy!

lazarevich changed discussion status to closed Jan 12

AImhotep

Jan 12

Thank you!

25% BF16: TBD <- dibs on that as I need to AWQ to 4 bits

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