Alibaba debuts free Qwen AI assistant app on app stores

Alibaba has launched Qwen, a free multipurpose AI assistant app, now available on Apple’s App Store and Google Play.

The app, developed by Alibaba Cloud, aims to provide AI-powered services for both professional and personal use.

Qwen integrates features such as mapping and shopping, and is based on Alibaba’s open-source model of the same name.

The company said the app experienced high user demand after launch, with some users reporting service delays on Weibo.

Qwen’s free access model contrasts with the subscription approaches of international competitors like OpenAI and Anthropic.

In China, the app will compete with ByteDance’s Doubao and Zhipu AI, which also offer basic services for free but charge for premium access.

Alibaba Cloud, the AI and cloud unit of Hangzhou-based Alibaba, has been registering Qwen trademarks in sectors including education, entertainment, social services, and finance.

🔗 Source: South China Morning Post

🧠 Food for thought

Implications, context, and why it matters.

Qwen’s regional focus could curb global reach

Availability of Qwen’s App Store and Google Play listings outside mainland China is still unclear 1. That will set whether this is a global push or a domestic launch.
Alibaba says an international version is planned 1 2. The current beta centers on China with maps, food delivery, plus ticket booking 2. These services need local partners and approvals.
Qwen competes with ByteDance’s Doubao and Zhipu AI in China 3. The near-term push looks local, not global.
Trademark filings in education, entertainment, and finance 3 hint at a China-first ecosystem before any overseas rollout.

Teams running multi Large Language Model (LLM) systems can use Qwen pricing to trim costs

Developers can pick from tiered models, from Qwen-Flash for speed and price to Qwen-Max for harder tasks 4. That gives room to match different use cases.
Alibaba Cloud offers a free API quota in the Singapore region, not Beijing 5. Teams can test integrations without upfront spend.
Third-party providers like Groq, a specialist in AI inference and model hosting, price Qwen3-32B at $0.29 per million input tokens 6. That can suit multi LLM routing that balances cost and performance.
Open-source Qwen models are available 4 7. Teams can deploy them on their own or call them through Alibaba Cloud Model Studio, its AI development platform.