Instructions to use aleada/Pixtral-12B-W4A16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use aleada/Pixtral-12B-W4A16 with Transformers:

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

pipe = pipeline("image-text-to-text", model="aleada/Pixtral-12B-W4A16")
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("aleada/Pixtral-12B-W4A16")
model = AutoModelForMultimodalLM.from_pretrained("aleada/Pixtral-12B-W4A16", 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 aleada/Pixtral-12B-W4A16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "aleada/Pixtral-12B-W4A16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aleada/Pixtral-12B-W4A16",
		"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/aleada/Pixtral-12B-W4A16

SGLang

How to use aleada/Pixtral-12B-W4A16 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 "aleada/Pixtral-12B-W4A16" \
    --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": "aleada/Pixtral-12B-W4A16",
		"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 "aleada/Pixtral-12B-W4A16" \
        --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": "aleada/Pixtral-12B-W4A16",
		"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 aleada/Pixtral-12B-W4A16 with Docker Model Runner:
```
docker model run hf.co/aleada/Pixtral-12B-W4A16
```

aleada commited on Jul 5

Commit

c62c19f

verified ·

1 Parent(s): c5ccbb3

Fix chat template: handle system message with multimodal (list) content

Browse files

The template did `"[INST]" + system_message`, assuming string content. OpenAI-compatible servers (vLLM / litellm) normalize content into the content-parts list form for multimodal models, so any request carrying a system message raised: can only concatenate str (not list) to str. Flatten system + assistant list content to their text chunks; the user branch was already list-aware.

Files changed (1) hide show

chat_template.jinja +38 -6

chat_template.jinja CHANGED Viewed

@@ -1,5 +1,30 @@
 {%- if messages[0]["role"] == "system" %}
-    {%- set system_message = messages[0]["content"] %}
     {%- set loop_messages = messages[1:] %}
 {%- else %}
     {%- set loop_messages = messages %}
@@ -12,9 +37,7 @@
     {%- endif %}
     {%- if message["role"] == "user" %}
         {%- if loop.last and system_message is defined %}
-            {{- "[INST]" + system_message + "
-" }}
         {%- else %}
             {{- "[INST]" }}
         {%- endif %}
@@ -33,8 +56,17 @@
         {%- endif %}
         {{- "[/INST]" }}
     {%- elif message["role"] == "assistant" %}
-        {{- message["content"] + eos_token}}
     {%- else %}
         {{- raise_exception("Only user and assistant roles are supported, with the exception of an initial optional system message!") }}
     {%- endif %}
-{%- endfor %}

+{#- Pixtral-12B chat template — patched 2026-07-05.
+   The original template did `"[INST]" + system_message` and
+   `message["content"] + eos_token`, assuming `content` is ALWAYS a
+   plain string. Pixtral is multimodal, so OpenAI-compatible servers
+   (vLLM, litellm, ...) normalize message content into the content-parts
+   LIST form (`[{"type": "text", "text": ...}]`) before the template
+   runs. A plain system/assistant message then arrives as a list and
+   `str + list` raises "can only concatenate str (not list) to str" —
+   any request carrying a `system` message 400'd.
+   Fix: flatten system + assistant content to their text chunks when
+   they arrive as a list. The user branch already handled both string
+   and list content (including [IMG]) correctly and is unchanged. #}
 {%- if messages[0]["role"] == "system" %}
+    {%- set system_raw = messages[0]["content"] %}
+    {%- if system_raw is string %}
+        {%- set system_message = system_raw %}
+    {%- else %}
+        {%- set sys_ns = namespace(text="") %}
+        {%- for chunk in system_raw %}
+            {%- if chunk["type"] == "text" %}
+                {%- set sys_ns.text = sys_ns.text + chunk["text"] %}
+            {%- endif %}
+        {%- endfor %}
+        {%- set system_message = sys_ns.text %}
+    {%- endif %}
     {%- set loop_messages = messages[1:] %}
 {%- else %}
     {%- set loop_messages = messages %}
     {%- endif %}
     {%- if message["role"] == "user" %}
         {%- if loop.last and system_message is defined %}
+            {{- "[INST]" + system_message + "\n\n" }}
         {%- else %}
             {{- "[INST]" }}
         {%- endif %}
         {%- endif %}
         {{- "[/INST]" }}
     {%- elif message["role"] == "assistant" %}
+        {%- if message["content"] is string %}
+            {{- message["content"] + eos_token }}
+        {%- else %}
+            {%- for chunk in message["content"] %}
+                {%- if chunk["type"] == "text" %}
+                    {{- chunk["text"] }}
+                {%- endif %}
+            {%- endfor %}
+            {{- eos_token }}
+        {%- endif %}
     {%- else %}
         {{- raise_exception("Only user and assistant roles are supported, with the exception of an initial optional system message!") }}
     {%- endif %}
+{%- endfor %}