The new model previews, the Qwen3.7-Max-Preview and the Qwen3.7-Plus-Preview, ranked 13th globally in text capabilities and 16th in vision capabilities, respectively, benchmark firm LM Arena said on Tuesday.

The versions overtake other Chinese AI labs in several areas, making them the top Chinese AI models of the moment, based on Arena’s rankings. The preview models still trail behind top US products, including Anthropic’s Claude, Google’s Gemini and OpenAI’s GPT models.

Alibaba’s proprietary Qwen3.7 preview models come a month after the company launched its previous flagship Qwen3.6 series. The Chinese AI and e-commerce giant, which owns the South China Morning Post, has not revealed details of the upcoming Qwen3.7 series models, though it has teased the launch of a major new model at the Alibaba Cloud Summit in the eastern city of Hangzhou on Wednesday.

Tech companies often release preview versions of their next-generation models on Arena, which ranks models based on user preferences, in order to collect data to optimise for the final iteration.

Alibaba is among Chinese AI developers that have recently moved away from fully open-sourcing their most powerful models, as commercialisation pressures increase, driven by the massive spending on computing power required for the products.

The company kept its previous Qwen3.6-Max and Qwen3.6-Plus models proprietary, which requires paid use through an API, or application programming interface.