Qwen-VL: A Frontier Large Vision-Language Model with Versatile Abilities

Abstract

We introduce the Qwen-VL series, a set of large-scale vision-language models designed to perceive and understand both text and images. Comprising Qwen-VL and Qwen-VL-Chat, these models exhibit remarkable performance in tasks like image captioning, question answering, visual localization, and flexible interaction. The evaluation covers a wide range of tasks including zero-shot captioning, visual or document visual question answering, and grounding. We demonstrate the Qwen-VL outperforms existing Large Vision Language Models (LVLMs). We present their architecture, training, capabilities, and performance, highlighting their contributions to advancing multimodal artificial intelligence. Code, demo and models are available at https://github.com/QwenLM/Qwen-VL.

Full paper on arXiv

Details

arXiv identifier

2308.12966

Published

2023-08-24

Authors

Jinze Bai, Shuai Bai, Shusheng Yang, Shijie Wang, Sinan Tan, Peng Wang, Junyang Lin, Chang Zhou, Jingren Zhou

Models That Cite This Paper

• Described byQwen3-VL-8B-Instruct
• Described byQwen2.5-VL-7B-Instruct
• Described byQwen3-VL-4B-Instruct
• Described byQwen3-VL-2B-Instruct
• Described byQwen2.5-VL-3B-Instruct
• Described byQwen3-VL-8B-Instruct-FP8
• Described byQwen2-VL-2B-Instruct
• Described byQwen2.5-VL-7B-Instruct-AWQ
• Described byQwen2-VL-7B-Instruct-AWQ
• Described byQwen3-VL-30B-A3B-Instruct-AWQ
• Described byQwen2-VL-7B-Instruct