- Alibaba Cloud summit is scheduled for May 20 as company teases new Qwen product.
- Preview versions of Qwen3.7 models have already gone live on testing platforms.
Alibaba is preparing to roll out the latest version of its AI (artificial intelligence) large model, hinting at the release of a "heavyweight" new product at the upcoming Alibaba Cloud summit.
The Qwen team released a teaser for the new product, describing it as "more versatile, more powerful, with depth and breadth," according to a Monday announcement.
Specific details about the new product have not yet been disclosed, with Qwen only stating that the answer will be officially revealed at the Alibaba Cloud summit on Wednesday.
However, based on the appearance of Qwen's official capybara mascot in the promotional poster, the upcoming release is likely the Qwen3.7 large model — the next major iteration of Qwen.
Notably, preview models bearing this version number have recently gone live across multiple testing platforms.
Qwen3.7-Max-Preview and Qwen3.7-Plus-Preview have already been made available in advance on the Qwen Chat and Arena AI platforms.
The anticipated launch follows Alibaba's open-source initiatives last month, continuing the rapid iteration pace of the Qwen large models.
The company released the Qwen3.6-35B-A3B model, which features 35 billion parameters, in mid-April.
The model activates only 3 billion parameters during each inference, effectively reducing computing costs and significantly improving operational efficiency.
It can also be seamlessly integrated into mainstream programming assistants, and its predecessor previously earned praise from Elon Musk.
The potential release of Qwen3.7 will further consolidate the company's competitiveness in the AI large model sector.
