Qwen/Qwen3-VL-8B-Instruct
1.1k20m
1---2license: apache-2.03pipeline_tag: image-text-to-text4library_name: transformers5---6<a href="https://chat.qwenlm.ai/" target="_blank" style="margin: 2px;">7 <img alt="Chat" src="https://img.shields.io/badge/%F0%9F%92%9C%EF%B8%8F%20Qwen%20Chat%20-536af5" style="display: inline-block; vertical-align: middle;"/>8</a>9 10 11# Qwen3-VL-8B-Instruct12 13 14Meet Qwen3-VL — the most powerful vision-language model in the Qwen series to date.15 16This generation delivers comprehensive upgrades across the board: superior text understanding & generation, deeper visual perception & reasoning, extended context length, enhanced spatial and video dynamics comprehension, and stronger agent interaction capabilities.17 18Available in Dense and MoE architectures that scale from edge to cloud, with Instruct and reasoning‑enhanced Thinking editions for flexible, on‑demand deployment.19 20 21#### Key Enhancements:22 23* **Visual Agent**: Operates PC/mobile GUIs—recognizes elements, understands functions, invokes tools, completes tasks.24 25* **Visual Coding Boost**: Generates Draw.io/HTML/CSS/JS from images/videos.26 27* **Advanced Spatial Perception**: Judges object positions, viewpoints, and occlusions; provides stronger 2D grounding and enables 3D grounding for spatial reasoning and embodied AI.28 29* **Long Context & Video Understanding**: Native 256K context, expandable to 1M; handles books and hours-long video with full recall and second-level indexing.30 31* **Enhanced Multimodal Reasoning**: Excels in STEM/Math—causal analysis and logical, evidence-based answers.32 33* **Upgraded Visual Recognition**: Broader, higher-quality pretraining is able to “recognize everything”—celebrities, anime, products, landmarks, flora/fauna, etc.34 35* **Expanded OCR**: Supports 32 languages (up from 19); robust in low light, blur, and tilt; better with rare/ancient characters and jargon; improved long-document structure parsing.36 37* **Text Understanding on par with pure LLMs**: Seamless text–vision fusion for lossless, unified comprehension.38 39 40#### Model Architecture Updates:41 42<p align="center">43 <img src="https://qianwen-res.oss-accelerate.aliyuncs.com/Qwen3-VL/qwen3vl_arc.jpg" width="80%"/>44<p>45 46 471. **Interleaved-MRoPE**: Full‑frequency allocation over time, width, and height via robust positional embeddings, enhancing long‑horizon video reasoning.48 492. **DeepStack**: Fuses multi‑level ViT features to capture fine‑grained details and sharpen image–text alignment.50 513. **Text–Timestamp Alignment:** Moves beyond T‑RoPE to precise, timestamp‑grounded event localization for stronger video temporal modeling.52 53This is the weight repository for Qwen3-VL-8B-Instruct.54 55 56---57 58## Model Performance59 60**Multimodal performance**61 6263 64**Pure text performance**6566 67## Quickstart68 69Below, we provide simple examples to show how to use Qwen3-VL with 🤖 ModelScope and 🤗 Transformers.70 71The code of Qwen3-VL has been in the latest Hugging Face transformers and we advise you to build from source with command:72```73pip install git+https://github.com/huggingface/transformers74# pip install transformers==4.57.0 # currently, V4.57.0 is not released75```76 77### Using 🤗 Transformers to Chat78 79Here we show a code snippet to show how to use the chat model with `transformers`:80 81```python82from transformers import Qwen3VLForConditionalGeneration, AutoProcessor83 84# default: Load the model on the available device(s)85model = Qwen3VLForConditionalGeneration.from_pretrained(86 "Qwen/Qwen3-VL-8B-Instruct", dtype="auto", device_map="auto"87)88 89# We recommend enabling flash_attention_2 for better acceleration and memory saving, especially in multi-image and video scenarios.90# model = Qwen3VLForConditionalGeneration.from_pretrained(91# "Qwen/Qwen3-VL-8B-Instruct",92# dtype=torch.bfloat16,93# attn_implementation="flash_attention_2",94# device_map="auto",95# )96 97processor = AutoProcessor.from_pretrained("Qwen/Qwen3-VL-8B-Instruct")98 99messages = [100 {101 "role": "user",102 "content": [103 {104 "type": "image",105 "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",106 },107 {"type": "text", "text": "Describe this image."},108 ],109 }110]111 112# Preparation for inference113inputs = processor.apply_chat_template(114 messages,115 tokenize=True,116 add_generation_prompt=True,117 return_dict=True,118 return_tensors="pt"119)120inputs = inputs.to(model.device)121 122# Inference: Generation of the output123generated_ids = model.generate(**inputs, max_new_tokens=128)124generated_ids_trimmed = [125 out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)126]127output_text = processor.batch_decode(128 generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False129)130print(output_text)131```132 133### Generation Hyperparameters134#### VL135```bash136export greedy='false'137export top_p=0.8138export top_k=20139export temperature=0.7140export repetition_penalty=1.0141export presence_penalty=1.5142export out_seq_length=16384143```144 145#### Text146```bash147export greedy='false'148export top_p=1.0149export top_k=40150export repetition_penalty=1.0151export presence_penalty=2.0152export temperature=1.0153export out_seq_length=32768154```155 156 157## Citation158 159If you find our work helpful, feel free to give us a cite.160 161```162@misc{qwen3technicalreport,163 title={Qwen3 Technical Report}, 164 author={Qwen Team},165 year={2025},166 eprint={2505.09388},167 archivePrefix={arXiv},168 primaryClass={cs.CL},169 url={https://arxiv.org/abs/2505.09388}, 170}171 172@article{Qwen2.5-VL,173 title={Qwen2.5-VL Technical Report},174 author={Bai, Shuai and Chen, Keqin and Liu, Xuejing and Wang, Jialin and Ge, Wenbin and Song, Sibo and Dang, Kai and Wang, Peng and Wang, Shijie and Tang, Jun and Zhong, Humen and Zhu, Yuanzhi and Yang, Mingkun and Li, Zhaohai and Wan, Jianqiang and Wang, Pengfei and Ding, Wei and Fu, Zheren and Xu, Yiheng and Ye, Jiabo and Zhang, Xi and Xie, Tianbao and Cheng, Zesen and Zhang, Hang and Yang, Zhibo and Xu, Haiyang and Lin, Junyang},175 journal={arXiv preprint arXiv:2502.13923},176 year={2025}177}178 179@article{Qwen2VL,180 title={Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution},181 author={Wang, Peng and Bai, Shuai and Tan, Sinan and Wang, Shijie and Fan, Zhihao and Bai, Jinze and Chen, Keqin and Liu, Xuejing and Wang, Jialin and Ge, Wenbin and Fan, Yang and Dang, Kai and Du, Mengfei and Ren, Xuancheng and Men, Rui and Liu, Dayiheng and Zhou, Chang and Zhou, Jingren and Lin, Junyang},182 journal={arXiv preprint arXiv:2409.12191},183 year={2024}184}185 186@article{Qwen-VL,187 title={Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond},188 author={Bai, Jinze and Bai, Shuai and Yang, Shusheng and Wang, Shijie and Tan, Sinan and Wang, Peng and Lin, Junyang and Zhou, Chang and Zhou, Jingren},189 journal={arXiv preprint arXiv:2308.12966},190 year={2023}191}192```