Every so often, Andreessen Horowitz’s (a16z) blog publishes updates to its report on the global generative AI market, covering both web and app usage based on Similarweb and Sensor Tower data. In this edition, the main players are the ones you would expect—ChatGPT and Gemini—but as you will see, there is also room for many other names worth knowing.
This is the list of the top 10 generative AI tools on the web, ranked by unique monthly visits:
And yes, just this once, we will mention number 11 as well: Spain’s own Freepik, recently rebranded as Magnific.
As you can see, on the web—and to no one’s surprise—ChatGPT remains the undisputed leader. No other service comes close to its traffic volume, although the gap with Gemini is beginning to narrow, as we explained in our overview of the general-purpose AI market.
In third place, Canva, repositioned as a full-fledged AI-powered conversational design platform, has pushed down China’s DeepSeek, which gained major visibility following its launch.
Alongside them are others such as Perplexity, which has become the preferred alternative for those looking for a “Google with AI”, and the ever-controversial Grok, the assistant integrated into X.
From positions 11 to 50, the ranking becomes much more diverse.
Among these generative AI tools, standout platforms include those focused on editing and creating visual content, such as Remove.bg, Pixelcut, and Photoroom, which specialize in enhancing images or automatically generating graphics. Also notable are tools centered on text generation and language assistance, such as NotebookLM (from Google) and Kimi, as well as AI-powered messaging alternatives such as SpicyChat AI, GigaChat, and Candy.ai.
Also noteworthy is the presence, in 13th place, of the relatively little-known Janitor.ai. This is a platform for creating and interacting with customized AI characters and chatbots, rather than a standalone AI model. It allows users to create and chat with virtual characters for entertainment, role-playing, and other creative projects.
In the audio and video generation space, names such as Suno and ElevenLabs appear, while Midjourney continues to hold a strong position in image generation and visual experiences.
China’s influence is clearly visible in the web ranking, where names such as Deepseek, Quark, Doubao, and Kimi appear. There is a twofold explanation for this. On the one hand, access to Western tools such as ChatGPT, Perplexity, and Claude is blocked or restricted in China, which encourages the development of local alternatives. On the other hand, in areas such as video generation, Chinese research teams have taken the lead, partly thanks to looser copyright regulation. This Team China category could also include Manus, although—pending the resolution of its legal status—it is already part of Meta’s structure.
The report also includes the Top 10 generative AI mobile apps by monthly active users:
The mobile ranking reflects a similar dynamic, although with some nuances. ChatGPT also tops this list, but in this segment two apps closely tied to the social media space appear: Capcut, the app that shares a parent company with TikTok and allows users to edit their short videos, and once again Canva, which over time has become an essential reference point for content creation specialists worldwide.
Among the top positions are apps that are not especially well known in our market, such as Doubao, Meituan, and Yandex, but which lead massive markets in Asia and Eastern Europe. For example, Doubao is the AI assistant developed—like Capcut and TikTok—by ByteDance, and it has positioned itself as the most popular chatbot in China. Meanwhile, Meituan operates as a super app that centralizes local services, dominating the food delivery sector while also covering hotel bookings and ticket purchases within a single ecosystem. Finally, Yandex is something like the Russian market’s equivalent of Google, offering everything from a powerful search engine and map services to transportation and eCommerce solutions, consolidating its role as a foundational technology pillar for millions of users outside the Western sphere.
Just below the top 10 is Remini. Available as an app for iOS and Android, Remini is an AI tool that improves the quality of photos and videos, allowing users to restore old images, sharpen blurry photos, reduce noise, enhance portraits, and even transform faces into different styles such as pop stars or runway models.
In fact, 12 of the 50 most-used mobile apps were developed in that country, although most of them primarily target international markets. Beyond those already mentioned, these include:
This prominence becomes even more pronounced when looking at the top 50 by web traffic, where names such as the following appear:
But the study does not focus only on the sector’s giants. Among the most interesting generative AI trends it identifies is the rise of so-called vibe coding, driven by the arrival of platforms that allow any user to build applications without needing to code, using natural language prompts. In this context, the person responsible for programming no longer creates code line by line, but instead “helps” an AI assistant generate, refine, and debug an application through conversation.
Genspark, Cursor, and especially Lovable are the most prominent examples in the vibe coding space. In the latter case, we are talking about a tool that allows users to build complete web applications (front-end and back-end) simply by chatting with the AI, without having to touch a single line of code.
Overall, the report leaves the impression that the generative AI market is no longer growing simply through the emergence of clones, but through differentiation. Users are gravitating toward tools that solve a specific need for them, whether conversational assistants, AI tools for creating images, video editors, or even— as with vibe coding—ways to program without technical knowledge. The battle among major assistants such as ChatGPT, Gemini, and Claude will largely define the immediate future, but at the same time there is ample room for specialized offerings that know how to find their audience.
Ultimately, this map of the world’s most-used applications is a snapshot of how AI is transforming everyday digital habits: what we search for, how we communicate, how we express ourselves, and even how we work. And if Andreessen Horowitz’s study reveals anything, it is that although the ecosystem is beginning to stabilize, there is still plenty of ground left to explore.
Image: Flux Schnell
Your email address will not be published. Required fields are marked *
Δ