Instructions to use Qwen/Qwen3.8-27B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen3.8-27B with Transformers:

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

pipe = pipeline("image-text-to-text", model="Qwen/Qwen3.8-27B")
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)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("Qwen/Qwen3.8-27B")
model = AutoModelForMultimodalLM.from_pretrained("Qwen/Qwen3.8-27B", 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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Qwen/Qwen3.8-27B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen3.8-27B"
# 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/Qwen3.8-27B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen3.8-27B

SGLang

How to use Qwen/Qwen3.8-27B 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/Qwen3.8-27B" \
    --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/Qwen3.8-27B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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/Qwen3.8-27B" \
        --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/Qwen3.8-27B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen3.8-27B with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen3.8-27B
```

Same stupid shilt like the qwen3.8-max, degenerates, forgets instructions.

#104

by Phil2Sat - opened Aug 16

Discussion

Phil2Sat

Aug 16

even if i try to promt it, using its big brother.
the give a shit on what the user wants, they forget instructions they dont care about what user asks for.
even the 3.8 max preview is useless every shit AI does the same. the all are degenerated lazy pigs.

maybe its because they areb based all on each other, the dont get things, thex fix A, but forget B,C,D,E,F,G. then you aask to fix them all.
they forget theri prompt trying to fix A-G, making new mistakes, after all the fail in A-Z.

i dont know, but its not size, not quant, its LAZY trainig. chatgpt does this, qwen does this, gemini does this.

i need more time jelling at the model do check the absolute basic logic than doing actual meanigful things, and model after model it gets worse.
i remeber times 2yrs ago, i had a task, gave it to ai, problem solved. thats the past.

actually ai is more human than ever, the more human brain degenerates the more ai degenerates, lobotomized illogical useless braindead shit.

Am i the only fighting against!?

clecho52

Aug 16

No one is forcing you to use it, works fine for me

clecho52

Aug 16

Actually even better, you can use your beloved long horizon task solving 2 year old models, im sure theyre still around

23Olivilo

Aug 16

try an other harness, i had most issues not onyl with the LLM more with the new way of useabilites of them, so think on this every situation get you angry , shows you for free a mastery you didnt get till now :)

darkmatter2222

Aug 16

Works fine for me in Claude code

Phil2Sat

Aug 16

•

edited Aug 16

And thats the problem, its just coding, coding works in claude code or loops but as univeral llm fail. everybody codes, then call it qwen code only usefull in combination with claude code or agentic loops.

thats exaclty the reason... goto chat.qwen.ai try 3.8 max stupid nonsense, same here.

we all are now DEVS silly DEVS no one can code a single line DEVS cause of claude code, yay it can do a silly webgame, or a silly app, but if it gets real complicated fail fail fail....

synthetic good, in real bad.

Phil2Sat changed discussion status to closed Aug 16

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