The Murky Waters of AI App Rankings
Attempting to rank AI applications raises fundamental questions about which metrics to consider and their relative importance. However, a review of numerous ranking lists reveals a puzzling phenomenon: the ranking of a single application can vary dramatically across different lists.
This inconsistency stems from two primary factors:
Divergent Ranking Methodologies
定焦One (Dingjiao One) discovered that some ranking lists aggregate data from both the app and web versions of the same application to derive a single ranking, while others separate them into distinct app and web rankings. Ranking criteria also vary considerably, with some relying on monthly active users (AI Product Ranking, Xsignal), others on daily active users (AIGCRank), and still others on download volume (New Rank). Certain lists incorporate two or three metrics, while others rely on a single criterion. This lack of uniformity makes direct comparisons challenging and highlights the need for a more standardized approach to evaluating AI app performance. The methodological differences significantly impact the final rankings, leading to varied and sometimes contradictory results across different platforms. Understanding these differences is crucial for interpreting the rankings accurately and avoiding misleading conclusions. The choice of metric, whether MAU, DAU, or download volume, reflects different priorities and perspectives on user engagement and app popularity. Some methodologies prioritize broader reach (MAU), while others focus on more frequent and consistent usage (DAU).
Data Discrepancies for the Same App and Metric
Even when evaluating the same application using the same metric, significant discrepancies exist across different ranking lists. For instance, both AI Product Ranking and Xsignal primarily use monthly active users to rank AI apps, but their data for individual apps differs substantially. As an example, AI Product Ranking reported a February MAU of 61.81 million for the DeepSeek app, while Xsignal reported 145.5077 million. These substantial variations in reported MAU figures for the same app on the same platform point to significant inconsistencies in data collection and reporting.
According to a veteran data analyst, the primary cause of these discrepancies lies in the varying data sources used by different ranking providers. He explained that AI ranking data is typically sourced in one of two ways: direct monitoring or third-party acquisition. Companies that directly monitor data are often data analytics firms that have expanded into AI app rankings following the surge in AI popularity. Other companies without in-house data capabilities purchase data from third-party providers and compile their own rankings. The reliance on third-party data introduces potential inaccuracies and makes it difficult to verify the reliability of the information. The analyst emphasized that the lack of standardized methodologies in the nascent AI industry further contributes to these data inconsistencies.
Reliance on third-party data introduces the risk of using non-primary sources and makes it difficult to assess data reliability. The analyst also pointed out that the nascent nature of the AI industry means that data analytics firms have yet to develop standardized methodologies for measuring AI app performance, leading to data inconsistencies. The discrepancies highlight the need for greater transparency and standardization in data collection practices within the AI app ranking ecosystem. Without a common framework for measuring app performance, it becomes difficult to compare and evaluate different apps objectively.
Another ranking provider admitted that their company purchases AI app data from third-party sources and struggled to reconcile the significant discrepancies across different providers during the compilation process. After careful consideration, they ultimately decided to abandon the commonly used MAU metric in favor of DAU for their rankings. This decision reflects the challenges in relying on inconsistent third-party data and the need to find alternative metrics that provide a more reliable representation of app performance. The switch to DAU suggests a prioritization of frequent user engagement over broader reach, potentially reflecting a different perspective on app value and user behavior.
Despite these discrepancies, a general consensus has emerged within the industry, according to multiple data analysts: MAU > DAU > Download Volume > Other in terms of metric importance. The preference for MAU as the primary metric reflects its ability to capture the overall reach and popularity of an app, while DAU provides insights into the frequency of use. Download volume, while important, is considered a less reliable indicator of long-term engagement. The relative importance of these metrics can vary depending on the specific goals and priorities of the ranking provider.
‘If an application consistently ranks highly across different lists, it’s generally safe to assume that it’s performing well,’ one analyst noted. Regarding specific data discrepancies, industry professionals advise against excessive scrutiny, emphasizing the importance of focusing on overall trends. While acknowledging the limitations of the data, experts recommend looking for consistent patterns and overall trends across multiple ranking lists to gain a more accurate understanding of app performance. Focusing on the big picture can help to mitigate the impact of individual data discrepancies and provide a more reliable assessment of the AI app landscape.
Based on these insights, 定焦One primarily relied on MAU data from various ranking lists for its analysis, averaging data across different lists when discrepancies arose. It is also important to note that many applications are available in both app and web versions, but app versions tend to indicate higher user engagement and proactive usage. Therefore, the analysis primarily focuses on app data, while also presenting some app data with significant fluctuations to provide a more comprehensive reference point. The approach of averaging data across multiple lists aims to minimize the impact of individual data inaccuracies and provide a more balanced and reliable assessment. The focus on app data reflects the understanding that app usage generally indicates higher levels of user engagement compared to web-based access.
Global Showdown: Popular App Types and the Rise of Educational Tools
Let’s first examine the top 20 global AI apps for a single month.
In January, ChatGPT maintained a commanding lead in the global AI app rankings, owing to its early entry and advanced technology. It was the first AI app to achieve widespread popularity, and it now enjoys a significant advantage in MAU and monthly downloads, with 349.41 million and 695 million, respectively, far exceeding second-place Quark’s 143.4 million and 5.9374 million. ChatGPT’s dominant position reflects its pioneering role in the AI app space and its ability to attract a large and engaged user base. Its massive MAU and download numbers underscore its widespread adoption and its continued relevance in the evolving AI landscape. The significant gap between ChatGPT and its competitors highlights the challenges that other apps face in trying to dethrone the industry leader.
Following closely behind were AI apps from major Chinese tech companies, with ByteDance’s Doubao and the rising star DeepSeek ranking third and fourth, respectively. While other AI apps have been fixtures on the list, DeepSeek’s entry was nothing short of explosive. Previously unknown, it surged to the top ranks of the list following the release of its DeepSeek-R1 model before the Spring Festival. The rapid rise of Chinese AI apps, particularly Doubao and DeepSeek, showcases the increasing competitiveness of the Chinese tech industry in the AI sector. DeepSeek’s impressive debut highlights the potential for new entrants to disrupt the market with innovative technologies and attract a significant user base in a short period of time.
Among the ‘AI Six Little Dragons,’ Moonshot AI’s Kimi ranked eighth, while Zhipu AI’s Zhipu Qingyan ranked eighteenth. Baidu’s Wen Xin Yi Yan, 360’s Nami Search, and iFlytek’s iFlytek Spark lagged behind in the rankings, all falling outside the top ten. The mixed performance of the ‘AI Six Little Dragons’ suggests varying levels of success in attracting and retaining users. The lower rankings of established players like Baidu, 360, and iFlytek indicate the challenges in competing with newer, more agile entrants in the rapidly evolving AI app market.
Two major events shaped the AI application market in February: the integration of DeepSeek into various domestic AI apps and the release of product updates by international companies in response to DeepSeek’s emergence. These developments had a significant impact on the month’s rankings. The integration of DeepSeek’s technology into other AI apps highlights the growing trend of collaboration and partnerships within the industry. The response from international companies to DeepSeek’s success demonstrates the increasing competitiveness of the global AI app market.
Notably, Tencent’s AI chatbot app Yuanbao and Anthropic’s Claude, a prominent US AI startup, jumped from outside the top 20 to tenth and sixteenth, respectively. Yuanbao benefited from the February 13 announcement that it had integrated DeepSeek, with AI Product Ranking reporting a MAU of 13.12 million for the month. Claude, on the other hand, launched a major update in February, adding its ‘most intelligent model to date,’ Claude 3.7 Sonnet, and introducing mixed reasoning capabilities for the first time, generating considerable market excitement. The rise of Yuanbao following its integration with DeepSeek underscores the benefits of strategic partnerships and technology sharing. Claude’s improved ranking highlights the importance of continuous innovation and product updates in maintaining competitiveness in the fast-paced AI app market.
In March, the market witnessed the debut of Manus, an AI Agent app hailed as ‘comparable to DeepSeek and shocking Silicon Valley,’ also developed by a Chinese startup team. However, due to its invitation-only access, most users were unable to experience the product, and it did not make the top 20. As a result, the AI app rankings saw little change in March, with Yuanbao climbing from tenth to fifth. Wen Xin Yi Yan also integrated DeepSeek in March, improving its ranking from seventeenth in February to thirteenth. The debut of Manus, despite its limited accessibility, generated significant buzz and highlights the potential for AI Agent apps to disrupt the market. The lack of significant changes in the overall rankings in March suggests a period of consolidation and stability following the earlier disruptions.
It is worth noting that several vertical-specific apps also made their way into the top ranks of various lists. For instance, Xsignal’s list showed that AI writing tool AI Creation Lion had a February MAU of 3.7237 million, a staggering 20,011.43% increase month-over-month. AIGCRank’s list showed that AI educational tools such as Homework Help’s Quick Right AI and Yuanfudao’s Xiaoyuan AI made their debut in the top 20 domestic apps thanks to high DAUs. The emergence of vertical-specific AI apps, such as AI writing tools and AI educational tools, demonstrates the growing diversification of the AI app market and the increasing demand for specialized solutions. The impressive growth of AI Creation Lion highlights the potential for niche apps to achieve rapid success by catering to specific user needs.
Overall, ChatGPT, Quark, Doubao, and DeepSeek dominated almost all rankings. Beyond this, several conclusions can be drawn:
First, in terms of app type, the overall rankings from January to March indicate the following order of popularity: Chatbots > AI Companions > AI Image Editors > AI Office Tools > AI Video Tools.
The top four apps are all in the chatbot category, followed by AI companions (AI emotional support) represented by Talkie AI and Hoshino, and AI image editing tools such as Remini. AI office tools such as DeepL and Notion AI rank relatively lower. The dominance of chatbots reflects their versatility and broad appeal, while the popularity of AI companions highlights the growing demand for emotional support and social connection. The lower ranking of AI office tools suggests that these apps may require further development to fully integrate into users’ daily workflows.
According to one industry professional, chatbots are positioned as general-purpose AI tools and naturally enjoy high penetration rates. Major tech companies are primarily focusing on this direction. AI companions, on the other hand, cater to the emotional companionship and social needs of today’s youth. As for AI image editing and office apps, they are closely related to people’s daily lives and work. The strategic focus of major tech companies on chatbots underscores their potential to become ubiquitous tools for a wide range of applications. The popularity of AI companions highlights the increasing importance of addressing emotional and social needs in the digital age. The relevance of AI image editing and office apps to daily life and work suggests opportunities for further innovation and growth in these categories.
In contrast, despite significant investment from companies such as ByteDance, Kuaishou, and MiniMax in the AI video space, consumer adoption has lagged behind other types of apps, primarily due to a higher barrier to entry. As for niche apps such as AI diagnosis and AI programming, they are still in the market cultivation stage and require further adoption. The slower adoption of AI video apps may be attributed to the complexity of video creation and editing, which requires specialized skills and knowledge. The limited adoption of niche apps like AI diagnosis and AI programming reflects the need for further development and user education to demonstrate their value and address specific user needs.
Each of the top five AI categories has produced at least one representative product.
For example, ChatGPT is the leader in the chatbot category and is experiencing stable growth, with March downloads far exceeding January. According to one industry professional, ChatGPT’s MAU and downloads are expected to continue growing given its current rate of technological iteration. The competition in the domestic chatbot space is more intense, with Quark, Doubao, and DeepSeek not showing a significant gap between them. In the AI companion category, Talkie AI, developed by Chinese company MiniMax specifically for overseas markets, has the highest MAU. ChatGPT’s continued success reflects its ability to maintain its competitive edge through ongoing innovation and technological advancements. The intense competition in the domestic chatbot market highlights the challenges for companies seeking to establish a dominant position. The success of Talkie AI in the AI companion category demonstrates the potential for companies to target specific markets and user needs.
Second, in terms of the competitive landscape between Chinese and international companies.
Multiple industry professionals believe that Chinese AI apps are developing rapidly, and the gap between Chinese and international apps is narrowing. In the top 20 apps for Q1, the number of Chinese and international apps is roughly equal, with Chinese apps occupying four of the top five spots and Tencent Yuanbao showing the fastest growth. The increasing competitiveness of Chinese AI apps highlights the rapid advancements in the Chinese tech industry and its growing influence on the global AI landscape. The strong performance of Chinese apps in the top rankings suggests that they are effectively addressing user needs and competing with established international players.
This is due in part to the increasing willingness of people to embrace AI and in part to the continued investment of major companies in AI app research and development, as well as the breakthrough of startups such as DeepSeek. The suspense in this global AI race continues to escalate. The increasing adoption of AI technology and the significant investments in R&D by both major companies and startups are driving the growth and innovation in the AI app market. The global AI race is becoming increasingly competitive, with companies from different countries vying for market share and technological leadership.
Domestic Melee: The Battle Heats Up, Yuanbao and Jimo Emerge as Key Wildcards
Let’s turn our attention to the domestic AI app landscape.
The top 10 AI apps in January were primarily dominated by three main groups: the ‘Big Tech’ group, including Alibaba’s Quark, ByteDance’s Doubao, Baidu’s Wen Xin Yi Yan, 360’s Nami AI Search, and iFlytek’s iFlytek Spark; the ‘AI Six Little Dragons,’ including Moonshot AI’s Kimi and Zhipu AI’s Zhipu Qingyan; and the ‘rising star’ DeepSeek. The dominance of these three groups highlights the competitive landscape of the domestic AI app market, with established tech giants, emerging startups, and specialized AI companies vying for market share.
Their primary focus is on the chatbot space, with the Big Tech group performing the strongest, with Quark and Doubao occupying the top two spots and DeepSeek ranking third. The strong performance of the Big Tech group in the chatbot space reflects their established user bases, extensive resources, and ability to leverage their existing platforms to promote their AI apps.
Two other apps in the top 10 do not belong to the chatbot category: ByteDance’s Maoxiang and MiniMax’s Hoshino, both of which are AI companion products. Their data across various metrics is very similar, and no clear winner has emerged. The presence of AI companion apps in the top 10 highlights the growing demand for emotional support and social connection in the digital age. The close competition between Maoxiang and Hoshino suggests a market opportunity for companies to differentiate their AI companion offerings and attract a larger user base.
The landscape changed significantly in February, with ByteDance’s AI video generation tool Jimo and Tencent’s Yuanbao making significant strides, ranking ninth and fifth, respectively. iFlytek Spark and Zhipu Qingyan disappeared from the top 10. The shifting rankings in February indicate the dynamic nature of the domestic AI app market, with new apps emerging and established players experiencing fluctuations in popularity. The rise of Jimo and Yuanbao highlights the potential for companies to disrupt the market with innovative features and strategic partnerships.
It is clear that the domestic chatbot space lacks differentiation, and its ranking is highly correlated with marketing efforts. The lack of differentiation in the domestic chatbot space suggests that companies need to focus on developing unique features and user experiences to stand out from the competition. The strong correlation between ranking and marketing efforts underscores the importance of effective marketing strategies in driving user adoption and gaining market share.
In March, Jimo advanced further, climbing from ninth to sixth in February, while its competitor, Kuaishou’s AI video tool Keling, remained outside the top 10. Jimo’s continued ascent reflects its growing popularity and its ability to attract users with its AI video generation capabilities. The contrasting performance of Jimo and Keling highlights the challenges in competing in the AI video space and the importance of developing innovative features and effective marketing strategies.
This may contradict the perceptions of many industry insiders. More than one industry professional believes that Keling’s text-to-video and image-to-video effects surpass those of OpenAI’s Sora. In March, Artificial Analysis, a globally recognized AI benchmark testing organization, released its latest global video generation ranking, with Keling 1.6pro (high-quality mode) topping the image-to-video category. The discrepancy between industry perceptions and user adoption suggests that technical superiority alone is not sufficient for success in the AI app market. Factors such as user experience, marketing, and accessibility also play a crucial role in driving user adoption and gaining market share.
Some industry professionals speculate that AI video generation tools have a higher barrier to entry, resulting in fewer users than other tools. Furthermore, many users primarily use the web version rather than the app version, so web data should also be considered when comparing the two. The higher barrier to entry for AI video generation tools may be attributed to the complexity of video creation and editing, which requires specialized skills and knowledge. The preference for web-based access among some users highlights the importance of considering both app and web data when evaluating the performance of AI video apps.
Xsignal’s list shows that Keling and Jimo’s web rankings have been close to each other for the past three months, with Jimo slightly ahead of Keling. Andreessen Horowitz (a16z), a world-renowned venture capital firm, released its 2025 Global 100 Generative AI Apps ranking, showing that Keling’s website had a monthly unique visitor count of 20, exceeding well-known overseas products such as Sora, Midjourney, and Runway. Jimo was not on the list. The contrasting data from different sources underscores the challenges in accurately assessing the performance of AI apps and the need to consider a variety of metrics and data points.
Whether Keling or Jimo enjoys greater market recognition may require a larger user base for these applications to draw a more accurate conclusion. The long-term success of Keling and Jimo will depend on their ability to attract a larger user base and demonstrate their value to a wider audience.
The ranking changes reveal that the chatbot space is the most competitive. In the first three months of this year, Quark, Doubao, and DeepSeek have consistently remained in the top tier of the domestic AI app rankings. Yuanbao is the biggest wildcard, not appearing on the list in January but surpassing Wen Xin Yi Yan and Nami AI Search in February to rank fifth, and then surpassing Kimi in March to rank fourth. The remaining apps have either remained stagnant or declined in ranking month by month. The consistent performance of Quark, Doubao, and DeepSeek highlights their established positions in the chatbot market. The rapid rise of Yuanbao demonstrates the potential for new entrants to disrupt the market with innovative features and strategic partnerships.
It should be noted that Yuanbao’s rapid ascent is largely due to its integration with DeepSeek. While Quark also integrated DeepSeek, its popularity is primarily due to its rich and convenient features, bringing together AI search, AI dialogue, and AI PPT, AI translation, AI image generation and other different needs. Yuanbao’s success underscores the benefits of strategic partnerships and technology integration. Quark’s popularity reflects its comprehensive feature set and its ability to cater to a wide range of user needs.
Looking at the quarterly battle, three key threads are becoming clear: First, multi-line layouts have become standard for top players. For example, ByteDance simultaneously bets on AI applications in three different directions: Doubao, Jimo and Maoxiang. MiniMax is involved in AI video and AI companions. The adoption of multi-line layouts by top players reflects the recognition of the need to diversify their AI app offerings and cater to a wider range of user needs.
Second, the war is intensifying.
Tencent Yuanbao was able to rise in February, in addition to eating DeepSeek’s dividends, but also used a large number of placements inside and outside the platform, and finally a great effort to create miracles. App Growing data shows that its estimated investment amount in the entire Q1 reached 1.7 billion, only tens of millions in January, and the investment amount directly reached 300 million in February, and in March it was close to 1.4 billion (the amount of investment is mainly calculated based on the number of advertising creatives and the list price of the APP in the corresponding placement, because most of the Yuanbao placement is in the Tencent channel, the actual expenditure may be lower than this amount). Under the burning money strategy of large factories, Kimi, who used to invest heavily, gradually reduced the intensity of placement, leading to a decline in ranking. The significant increase in marketing investment by Tencent Yuanbao demonstrates the intensifying competition in the AI app market and the willingness of companies to invest heavily in user acquisition. The decline in Kimi’s ranking following a reduction in marketing spend highlights the importance of sustained marketing efforts in maintaining market share.
App Growing data shows that the top ten AI applications with the most active MAU in China (of which DeepSeek and Quark have no relevant data for the time being) are also maintaining high placement, and their total monthly placement amount in January-March reached 400 million, 500 million, and 1.6 billion respectively. That is to say, the top hot AI applications in China spent 2.5 billion on advertising in one quarter. The significant advertising expenditure by the top AI apps in China underscores the intense competition for user attention and market share.
Third, technical advantages are not completely positively correlated with market performance, and Keling and Jimo’s recognition at the practitioner end is in the opposite state to the monthly active users and downloads at the user end. The divergence between technical recognition and market performance highlights the importance of factors beyond technical superiority in driving user adoption and gaining market share.
The AI application has not yet completely won, and in this arena, yesterday’s disruptor may become tomorrow’s goalkeeper, and whether it can balance technological innovation and market insight tests every player. The AI app market remains highly dynamic and unpredictable, with established players facing constant challenges from new entrants. The ability to balance technological innovation with market understanding and user needs is crucial for success in this evolving landscape.