Instructions to use Qwen/Qwen3-Coder-30B-A3B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen3-Coder-30B-A3B-Instruct with Transformers:

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

pipe = pipeline("text-generation", model="Qwen/Qwen3-Coder-30B-A3B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-Coder-30B-A3B-Instruct")
model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-Coder-30B-A3B-Instruct")
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 Qwen/Qwen3-Coder-30B-A3B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen3-Coder-30B-A3B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen3-Coder-30B-A3B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen3-Coder-30B-A3B-Instruct

SGLang

How to use Qwen/Qwen3-Coder-30B-A3B-Instruct 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 "Qwen/Qwen3-Coder-30B-A3B-Instruct" \
    --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": "Qwen/Qwen3-Coder-30B-A3B-Instruct",
		"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 "Qwen/Qwen3-Coder-30B-A3B-Instruct" \
        --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": "Qwen/Qwen3-Coder-30B-A3B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen3-Coder-30B-A3B-Instruct with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen3-Coder-30B-A3B-Instruct
```

Treating _ as * (tokenizer error?)

#10

by EmilPi - opened Aug 2, 2025

Discussion

EmilPi

Aug 2, 2025

I tested two versions of AWQ quants

https://huggingface.co/QuantTrio/Qwen3-Coder-30B-A3B-Instruct-AWQ
https://huggingface.co/stelterlab/Qwen3-Coder-30B-A3B-Instruct-AWQ
and two levels of unsloth GGUF quants https://huggingface.co/unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF
Q4_K_M
Q5_K_M
all having same type of output when reviewing Python code with init and other magic methods:

Code Review Summary 
Major Issues: 
     Syntax Errors: **eq** should be __eq__, **init** should be __init__, **str** should be __str__, and **name** should be __name__
...

I use recommended generation config.

...
--temp 0.7 --top-k 20 --min-p 0.0 --top-p 0.8 --repeat-penalty 1.05 \
...

set in both llama-server and Open WebUI.

EmilPi

Aug 3, 2025

I just checked with this original model weights. It still confuses _ and * tokens.

ELigoP

Aug 5, 2025

Hello, is there any way to fix this / work in progress? This makes model unusable for general python usage.

stev236

Aug 17, 2025

Are you still having this problem?
I'm using it with vLLM with FP8 quantization and not seeing this error. I asked the model if it can differentiates between '' and '**' and it said yes it can differentiate between '' and '**'.
Therefore the problem is probably not the with the model.

ELigoP

Aug 18, 2025

I tested with original fp16 weights (this repo) and mentioned AWQ weights (links above). I don't expect anything to have changed since I noticed this 15 days passed, and only chat template / tokenizer config was updated 11 days before, looks nothing related to this.

@stev236 the test should be like this:

---QUERY---
Review this Python code:

class A:
    def __init__(self, a):
        self.a = a
    def echo(self):
        print(self.a)
if __name__ == '__main__':
    a = A(1)
    A.echo()

---END QUERY---

EmilPi

Aug 18, 2025

I think I know the cause of issue:

This is Open WebUI issue, not of the Qwen model. Sorry to create this thread.

EmilPi changed discussion status to closed Aug 18, 2025

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