Qwen Unleashes a Flurry of Updates—Can It Challenge Doubao?

Alibaba’s Qwen AI has been making waves lately with a series of rapid updates. On August 7, it launched several new features including a thinking research module, scheduled tasks, an office assistant, voice calls, and a smart agent plaza. The office assistant stands out—it can operate computers and browsers, call skills, and deliver documents like PPTs, Word files, and Excel sheets. The smart agent plaza, meanwhile, integrates more daily life services into the app.

Three days later, on August 10, Qwen launched an open platform, with third-party services like SF Express, Ziru (a rental platform), and Hello Bike (car rental) being gradually integrated. The same day, its pricing plan was revealed: professional memberships for the office assistant are tiered, with monthly subscription prices ranging from 19 to 128 RMB. Video generation is charged separately by usage quota, with five tiers available.

QuestMobile data sheds light on Qwen’s growth trajectory. Last November, Qwen ranked sixth in monthly active users (MAU) among AI native apps. By March this year, its MAU reached 166 million, jumping to second place. However, growth slowed in June, with MAU only increasing to 167 million—while Doubao, the current leader, hit 382 million MAU in the same period. This gap is pushing Qwen to find new growth drivers.

To strengthen its core capabilities, Qwen has upgraded its models. On August 3, Alibaba released the Qwen3.8-Max, a flagship model with 2.4 trillion parameters and support for up to 1 million token context, enhancing programming, office, and long-cycle task capabilities. The image generation model Qwen-Image-3.0 and video generation model Wan3.0 (in public beta) have also been added to improve content creation abilities.

Qwen’s strategy is to evolve from a question-answering tool to a multi-functional AI application. Its office assistant connects to notes, calendars, and emails, while the open platform allows third-party services to join, enabling users to book SF Express deliveries, search Ziru listings, or get Hello Bike rental recommendations directly within Qwen.

Comparing with Doubao, the two apps take different paths. Doubao, backed by ByteDance’s mature traffic distribution and user growth system, focuses on lowering the barrier to AI use for ordinary users. Qwen, on the other hand, leverages Alibaba’s ecosystem—including models, cloud computing power, and business lines like Taobao and Alipay—to create a cohesive experience.

However, Qwen faces challenges. Alibaba’s resources are scattered across various apps (Taobao for shopping, Alipay for payments, Gaode for navigation), and users are accustomed to using specific apps for specific tasks. Qwen needs to demonstrate that its integrated services are more efficient than traditional apps—for example, helping users compare rental listings with commute data from Gaode.

On the commercialization front, Qwen’s pricing model focuses on task-specific services. Basic Q&A remains free, while paid services include tiered office memberships and video generation by quota. This approach aims to attract users with free features and generate revenue from professional, high-demand services.

Sources

• AIX Finance via Sina Finance (https://finance.sina.com.cn/stock/t/2026-08-12/doc-inimzqcr3701403.shtml)