2026 AI Application Semiannual Report: Doubao leads, Top 10 stable

```

On July 14, QuestMobile released the "2026 First Half AI Application Market Development Insight Report".

Data shows that as of May 2026, there is a clear differentiation in the penetration rates of the domestic AI application market across multiple terminals. Among them, AI-native apps reached a monthly active user scale (MAU) of 499 million, a year-on-year increase of 85.4%, becoming the fastest-growing channel.

In the past one to two years, large model vendors have gone through an intense "user acquisition battle," competing for users’ mobile home screens through massive traffic investment and pre-installation collaborations.

According to the latest data, the competition pattern among top players has stabilized after this round of traffic war. Doubao, Qianwen, and DeepSeek make up the current first tier of apps with over 100 million monthly actives, with MAUs of 382 million, 167 million, and 130 million, respectively.

Among them, Doubao, which started earlier and invested more in marketing, maintains a lead in absolute active user numbers; Qianwen’s year-on-year growth reached 5792.9%, demonstrating strong base expansion capacity; DeepSeek firmly holds third place thanks to its technical performance and community-drive.

Data indicates that the industry growth model relying purely on large-scale traffic investment for user acquisition is changing, with a trend towards user concentration among a few top apps becoming established.

Specifically, the report compiles user scale data for five types of AI application channels, reflecting a substantial shift in user interaction habits.

Aside from AI-native apps, terminal-manufacturer AI apps (such as AI assistants pre-installed in mobile systems) have the widest coverage, with MAU reaching 755 million, up 14.0% year-on-year; AI application plugins have a MAU of 644 million, down 7.9% year-on-year.

On the PC side, web-based AI apps have a MAU of 172 million, a sharp decrease of 22.8% year-on-year; while the PC client AI apps, though smaller in base, have a MAU of 18 million, up 20.1% year-on-year.

This comparison shows that early AI tool experiences relied heavily on PC web interfaces or existed as third-party software plugins, whereas now, independently packaged AI-native apps and system-level AI functions integrated by phone manufacturers have become mainstream. The decline in PC web-side activity confirms that AI tools are shifting from low-frequency "trial" scenarios to mobile-first daily use cases.

After the initial user base expansion, activity and usage depth have become the core metrics for measuring business health at this stage.

Data shows that as of May, there are significant differences in user stickiness across five types of AI apps, with AI-native apps leading at 92.7 times, followed by AI application plugins at 60.9 times, terminal-manufacturer AI apps at 51.4 times, PC client and PC web-side at 26.9 and 25.3 times, respectively.

Comparing the data, it’s clear that although AI apps pre-installed by phone manufacturers achieve 755 million monthly active users through "pre-installation" and underlying system privileges, their monthly per capita usage frequency is significantly lower than AI-native apps that require users to download and install.

This shows that while terminal pre-installed AI significantly lowers the threshold for user access to models, it has not yet demonstrated stronger stickiness in actual demand scenarios or deeper user retention compared to independent AI-native apps.

Overall, as AI-native app MAUs approach 500 million, the domestic AI application market has passed the pioneering phase driven by concept-based traffic attraction. Top vendors have not only widened the gap in MAUs compared to mid-level players, but the focus of competition has shifted clearly to retention metrics such as per capita usage frequency.

With infrastructure and underlying large model capabilities becoming increasingly similar, the next phase of market performance will directly depend on each company’s ability to analyze specific business scenarios and their engineering implementation level.

Risk Warning and DisclaimerThe market has risks, investment should be made cautiously. This article does not constitute personal investment advice and does not consider individual users’ special investment objectives, financial status, or needs. Users should consider whether any opinions, views, or conclusions in this article are suitable for their particular situation. Investing based on this article is at your own risk. ```