Instructions to use Qwen/Qwen3.5-35B-A3B-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen3.5-35B-A3B-FP8 with Transformers:

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

pipe = pipeline("image-text-to-text", model="Qwen/Qwen3.5-35B-A3B-FP8")
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("Qwen/Qwen3.5-35B-A3B-FP8")
model = AutoModelForMultimodalLM.from_pretrained("Qwen/Qwen3.5-35B-A3B-FP8")
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 Qwen/Qwen3.5-35B-A3B-FP8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen3.5-35B-A3B-FP8"
# 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/Qwen3.5-35B-A3B-FP8",
		"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/Qwen3.5-35B-A3B-FP8

SGLang

How to use Qwen/Qwen3.5-35B-A3B-FP8 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/Qwen3.5-35B-A3B-FP8" \
    --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/Qwen3.5-35B-A3B-FP8",
		"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/Qwen3.5-35B-A3B-FP8" \
        --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/Qwen3.5-35B-A3B-FP8",
		"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/Qwen3.5-35B-A3B-FP8 with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen3.5-35B-A3B-FP8
```

vllm nighlty and "No user query found in messages"

by dnum-ia-unistra - opened Mar 2

Discussion

dnum-ia-unistra

Mar 2

Hi,

I deployed that FP8 version with the latest nightly vllm (lmcache/vllm-openai:nightly-2026-03-02).
I use "qwen3_xml" as tool parser and i use the instruct mode with --default-chat-template-kwargs '{"enable_thinking": false}'

And i saw the following errors sometimes in the log file :

(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311] Error in preprocessing prompt inputs
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311] Traceback (most recent call last):
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/vllm/renderers/hf.py", line 472, in safe_apply_chat_template
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     return tokenizer.apply_chat_template(
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/transformers/tokenization_utils_base.py", line 1667, in apply_chat_template
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     rendered_chat, generation_indices = render_jinja_template(
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]                                         ^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/transformers/utils/chat_template_utils.py", line 539, in render_jinja_template
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     rendered_chat = compiled_template.render(
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]                     ^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/jinja2/environment.py", line 1295, in render
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     self.environment.handle_exception()
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/jinja2/environment.py", line 942, in handle_exception
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     raise rewrite_traceback_stack(source=source)
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "<template>", line 79, in top-level template code
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/jinja2/sandbox.py", line 401, in call
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     return __context.call(__obj, *args, **kwargs)
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/transformers/utils/chat_template_utils.py", line 447, in raise_exception
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     raise jinja2.exceptions.TemplateError(message)
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311] jinja2.exceptions.TemplateError: No user query found in messages.
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311] The above exception was the direct cause of the following exception:
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311] Traceback (most recent call last):
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/vllm/entrypoints/openai/chat_completion/serving.py", line 295, in render_chat_request
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     conversation, engine_prompts = await self._preprocess_chat(
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/vllm/entrypoints/openai/engine/serving.py", line 995, in _preprocess_chat
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     (conversation,), (engine_prompt,) = await renderer.render_chat_async(
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]                                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/vllm/renderers/base.py", line 755, in render_chat_async
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     for conv, prompt in await asyncio.gather(*rendered):
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/vllm/renderers/hf.py", line 694, in render_messages_async
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     prompt_raw = safe_apply_chat_template(
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]                  ^^^^^^^^^^^^^^^^^^^^^^^^^
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]   File "/opt/venv/lib/python3.12/site-packages/vllm/renderers/hf.py", line 487, in safe_apply_chat_template
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311]     raise ValueError(str(e)) from e
(APIServer pid=1) ERROR 03-02 14:02:34 [serving.py:311] ValueError: No user query found in messages.

In the latest lm studio, i saw i the changelog : "Fixed Qwen3.5 RAG jinja rendering bug: "No user query found in messages"
https://lmstudio.ai/changelog/lmstudio-v0.4.6

How can we fix it with vLLM ?

Thanks

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