Alibaba's Qwen Ai Chatbot Boasts 10 Million Downloads In Its First Week - Here's What It Offers

1 hour ago

Follow ZDNET: Add america arsenic a preferred source on Google.

ZDNET's cardinal takeaways

Qwen is now nan fastest-growing AI instrumentality successful history.
It debuted earlier this period successful nationalist beta.
Alibaba says it plans to grow nan app's functionality.

Chinese tech elephantine Alibaba's Qwen AI adjunct reached 10 cardinal downloads wrong nan first week pursuing its motorboat arsenic a nationalist beta app earlier this month, nan institution announced Monday.

That intends Qwen AI is increasing faster than its well-established rivals. For reference, ChatGPT reportedly reached a cardinal users wrong nan first 5 days pursuing its nationalist debut conscionable nether 3 years ago, which was viewed arsenic historical astatine nan time. ChatGPT remains nan astir celebrated chatbot by respective measures, but competitors for illustration Microsoft Copilot and Google's Gemini have precocious been gaining ground.

Also: The champion AI chatbots of 2025: I tested ChatGPT, Copilot, and others to find nan apical devices now

ChatGPT isn't disposable successful mainland China, nor are different starring AI devices built by US-based companies, specified arsenic Anthropic's Claude. Models for illustration Qwen, which are built by Chinese companies, person exclusive entree to nan country's organization of astir 1.4 cardinal people, astir 17% of nan world population.

What gives nan exemplary its edge?

The Nov. 17 beta motorboat of nan Qwen app -- disposable successful nan Chinese type of Apple's App Store arsenic Qwen Chat -- "marked Alibaba's astir important measurement yet into nan user AI market, aiming to toggle shape its cutting-edge foundational AI exemplary capabilities into real-life applications and tools," nan institution wrote successful its announcement. It added that nan instrumentality "is built to go nan AI-powered gateway to regular life." The app is presently only disposable for users successful mainland China.

The Qwen app specializes successful "deep research" -- an manufacture building describing an AI tool's expertise to hunt crossed a swathe of online resources and synthesize analyzable accusation -- arsenic good as vibe coding, AI-powered camera features, descent platform generation, and different functions, according to nan company.

Also: A caller Chinese AI exemplary claims to outperform GPT-5 and Sonnet 4.5 - and it's free

It's powered by nan Qwen3 family of models, which Alibaba introduced successful April. The institution said that nan largest of those, Qwen3-235B-A22B, is simply a mixture-of-experts (MoE) exemplary that exhibits skills successful coding, math, and different domains exceeding aliases comparable to different starring models, including OpenAI's o1, DeepSeek's R1, and Gemini 2.5 Pro.

Alibaba wrote successful its announcement that it plans to grow Qwen's repertoire, integrating "core manner and productivity services" specified arsenic nutrient delivery, "health guidance," recreation booking, and e-commerce.

"This ecosystem attack will empower nan adjunct to grip a wide array of real-world tasks," nan institution wrote, "equipping it arsenic a proactive, tin life and activity partner."

Also: Google's Antigravity puts coding productivity earlier AI hype - and nan consequence is astonishing

Moonshot, different Chinese tech developer, made headlines earlier this period pinch nan merchandise of its Kimi K2 Thinking model, which, according to nan company, outperforms OpenAI's GPT-5 and Anthropic's Claude Sonnet 4.5 connected cardinal testing benchmarks, including Humanity's Last Exam.

Government ties

Alibaba is simply a publically traded company, but arsenic Bloomberg reported early past year, entities owned by nan Chinese authorities declare astatine slightest partial ownership of much than a twelve of its business units.

A Reporters Without Borders investigation published successful September recovered that 3 starring AI models developed by Chinese labs, including Qwen, make outputs that "strictly align pinch Beijing's charismatic narratives, particularly erstwhile it comes to nan integrity of China's governmental system, its ideology and territorial claims." Models from DeepSeek, different Chinese AI lab, showed akin signs of censorship erstwhile they were released successful January.