Amid the buzz surrounding DeepSeek, domestic AI rival Alibaba Cloud’s (NYSE:BABA) own Qwen team released a new family of artificial intelligence models, Qwen2.5-VL, capable of performing a number of text and image analysis tasks.

Qwen is a series of large language models (LLMs) independently developed by Alibaba Cloud.

The new Qwen2.5-VL range can parse files, analyze videos, count objects in images and control computers – capabilities similar to OpenAI’s new Operator model.

Qwen said the top Qwen2.5-VL model fares better than Microsoft-backed (MSFT) OpenAI’s GPT-4o, Amazon-backed (AMZN) Anthropic’s Claude 3.5 Sonnet and Google’s (GOOG) (GOOGL) Gemini 2.0 Flash in math, document analysis, video analysis and question-answering evaluations.

The new LLMs are available to test in Alibaba’s Qwen Chat app and on the AI developer platform Hugging Face.

Qwen’s open-source AI LLMs are already some of the leading AI models in China, and its recent iterations have mostly been deployed for developers and business customers such as automakers, banks, video game makers and retailers as part of product development and shaping customer experiences.

Qwen2.5-VL can also interact with software on mobile devices and PCs. A post on X by Philipp Schmid from Hugging Face, showed the model launching and booking a flight on the Booking.com app for Android.

Other major AI models from Chinese players include: Doubao 1.5 Pro from TikTok parent ByteDance (BDNCE), Baidu's (BIDU) Ernie Bot 4.0 and Kimi k1.5 from AI startup Moonshot.