Instructions to use tencent/Youtu-VL-4B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tencent/Youtu-VL-4B-Instruct with Transformers:

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

pipe = pipeline("image-text-to-text", model="tencent/Youtu-VL-4B-Instruct", trust_remote_code=True)
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 AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("tencent/Youtu-VL-4B-Instruct", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use tencent/Youtu-VL-4B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tencent/Youtu-VL-4B-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": "tencent/Youtu-VL-4B-Instruct",
		"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/tencent/Youtu-VL-4B-Instruct

SGLang

How to use tencent/Youtu-VL-4B-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 "tencent/Youtu-VL-4B-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": "tencent/Youtu-VL-4B-Instruct",
		"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 "tencent/Youtu-VL-4B-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": "tencent/Youtu-VL-4B-Instruct",
		"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 tencent/Youtu-VL-4B-Instruct with Docker Model Runner:
```
docker model run hf.co/tencent/Youtu-VL-4B-Instruct
```

Add library_name and project page link

by nielsr HF Staff - opened Jan 29

base: refs/heads/main

←

from: refs/pr/3

Discussion Files changed

+10

-7

Files changed (1) hide show

README.md +10 -7

README.md CHANGED Viewed

@@ -1,15 +1,17 @@
 ---
 license: other
 license_name: youtu-vl
-extra_gated_eu_disallowed: true
 license_link: https://huggingface.co/tencent/Youtu-VL-4B-Instruct/blob/main/LICENSE.txt
 pipeline_tag: image-text-to-text
 ---
 <div align="center">
 # <img src="assets/youtu-vl-logo.png" alt="Youtu-VL Logo" height="100px">
-[📃 License](LICENSE.txt) • [💻 Code](https://github.com/TencentCloudADP/youtu-vl) • [📑 Technical Report](https://arxiv.org/abs/2601.19798) • [📊 Benchmarks](#benchmarks) • [🚀 Getting Started](#quickstart)
 </div>
 ## 🎯 Introduction
@@ -23,7 +25,7 @@ pipeline_tag: image-text-to-text
   - **Promising Performance with High Efficiency**: Despite its compact 4B-parameter architecture, the model achieves competitive results across a wide range of general multimodal tasks, including general visual question answering (VQA), multimodal reasoning and mathematics, optical character recognition (OCR), multi-image and real-world understanding, hallucination evaluation, and GUI agent tasks.
-  <p align="center">
       <img src="assets/youtu-vl-overview.png" width="90%"/>
   <p>
@@ -40,7 +42,7 @@ pipeline_tag: image-text-to-text
   - **Vision-Centric Prediction with a Standard Architecture (no task-specific modules)**: Youtu-VL treats image and text tokens with equivalent autoregressive status, empowering it to perform vision-centric tasks for both dense vision prediction (e.g., segmentation, depth) and text-based prediction (e.g., grounding, detection) within a standard VLM architecture, eliminating the need for task-specific additions. This design yields a versitile general-purpose VLM, allowing a single model to flexibly accommodate a wide range of vision-centric and vsion-language requirements.
-  <p align="center">
       <img src="assets/architecture.png" width="90%"/>
   <p>
@@ -49,7 +51,7 @@ pipeline_tag: image-text-to-text
 ### Vision-Centric Tasks
-  <p align="center">
       <img src="assets/vision-centric-performance.png" width="90%"/>
   <p>
@@ -57,7 +59,7 @@ pipeline_tag: image-text-to-text
 ### General Multimodal Tasks
-  <p align="center">
       <img src="assets/general-multimodal-performance.png" width="90%"/>
   <p>
@@ -122,7 +124,8 @@ outputs = processor.batch_decode(
     generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
 )
 generated_text = outputs[0]
-print(f"Youtu-VL output:\n{generated_text}")
 ```
 ## 🎉 Citation

 ---
 license: other
 license_name: youtu-vl
 license_link: https://huggingface.co/tencent/Youtu-VL-4B-Instruct/blob/main/LICENSE.txt
 pipeline_tag: image-text-to-text
+extra_gated_eu_disallowed: true
+library_name: transformers
 ---
 <div align="center">
 # <img src="assets/youtu-vl-logo.png" alt="Youtu-VL Logo" height="100px">
+[🏠 Project Page](https://youtu-tip.com/#llm) • [📃 License](LICENSE.txt) • [💻 Code](https://github.com/TencentCloudADP/youtu-vl) • [📑 Technical Report](https://arxiv.org/abs/2601.19798) • [📊 Benchmarks](#benchmarks) • [🚀 Getting Started](#quickstart)
 </div>
 ## 🎯 Introduction
   - **Promising Performance with High Efficiency**: Despite its compact 4B-parameter architecture, the model achieves competitive results across a wide range of general multimodal tasks, including general visual question answering (VQA), multimodal reasoning and mathematics, optical character recognition (OCR), multi-image and real-world understanding, hallucination evaluation, and GUI agent tasks.
+  <p align="center\">
       <img src="assets/youtu-vl-overview.png" width="90%"/>
   <p>
   - **Vision-Centric Prediction with a Standard Architecture (no task-specific modules)**: Youtu-VL treats image and text tokens with equivalent autoregressive status, empowering it to perform vision-centric tasks for both dense vision prediction (e.g., segmentation, depth) and text-based prediction (e.g., grounding, detection) within a standard VLM architecture, eliminating the need for task-specific additions. This design yields a versitile general-purpose VLM, allowing a single model to flexibly accommodate a wide range of vision-centric and vsion-language requirements.
+  <p align="center\">
       <img src="assets/architecture.png" width="90%"/>
   <p>
 ### Vision-Centric Tasks
+  <p align="center\">
       <img src="assets/vision-centric-performance.png" width="90%"/>
   <p>
 ### General Multimodal Tasks
+  <p align="center\">
       <img src="assets/general-multimodal-performance.png" width="90%"/>
   <p>
     generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
 )
 generated_text = outputs[0]
+print(f"Youtu-VL output:
+{generated_text}")
 ```
 ## 🎉 Citation