With a wild 2 million in 6 days, why is this AI app exploding all over the internet?

"LingGuang”, completely on fire.

Last week, Ant Group released its all-modal universalAI assistantAura, perhaps, is one of the biggest surprises in the AI app track at the end of 2025.

LingGuang's downloadsThe volume showed an accelerated growth trend, and on the first day of the launch, the download volume of Aura surpassed200,000Day 4 BreakthroughOne million.The latest figures show that the number of downloads of Aura has already exceeded2 millionStabilizedApp Store China free list sixth, free tools list firstThe location.

This growth rate surpasses a number of global phenomenal AI apps and far exceeds Aura's original internal plan to break 200,000 downloads by the end of the year. According to Appfigures, a mobile application analyzer, ChatGPT had 606,000 downloads in the first week of its launch, and it took five days for Sora2 to break a million, a process that Aura compressed to four days.

▲Aurora App surges to 2 million downloads in 6 days

Moreover, LingGuang is not an AI app that you can “experience and then leave behind”. The “LingGuang Flash App” feature, which allows everyone to rub their hands on the app, became a breaking point and was once used to the point of downtime.Only 4 days after launch, Aura had 8 rounds of emergency scaling due to high loadsThe main focus of the program is the “Aura Flash Application”.

Users no longer hold the curiosity to taste, but use it to solve practical problems. The average number of modification rounds of the flash application reaches 6 rounds, and some people even modified more than 100 rounds in 2 hours, which fully proves the strong demand for “practical AI”.

In other words, the achievement of breaking 2 million downloads in 6 days is not a coincidence, but a vote of approval from users with real behavior.

First, 2 million users “use explosion” aura, set off the hand rubbing application craze

Since its launch last Tuesday, Aura has already accumulated a lot of word-of-mouth quickly on the user side. Many users have shared their first-hand experience on social media platforms, giving a lot of genuine feedback on everything from speed, ease of use, effectiveness to product concept.

Semi-applicationPerhaps the most hotly discussed feature after the release of LingGuang, which set off a wave of hand rubbing applications. After opening the social platform and searching for “LingGuang”, you can see that many users have shared cases related to the flash application.From exclusive training tools, to to-do list applets and gas car trolley money-saving calculatorsetc., Aura has created a number of useful applications for users that can boost efficiency in real life.

There are also netizens who have realized a number of fun applications with the help of flash apps, such asBoss Talk Translator, Tutoring Homework Cyber Merit Box, and Baby Walker DrawerAnd so on. These interesting but especially “grounded” use of the scene, so that people can intuitively feel that the aura really began to melt into the daily life of everyone, whether it is to enhance the efficiency or the whole point of creativity, it can help.

Liu Fei, a well-known business technology podcast anchor, shared that he tried to make a wallpaper generator and a small app for Japanese language learning in the Aura Flash App feature for a few minutes. Lingguang also completed a variety of underlying support on its own, such as networking services, the experience is very smooth.

Liu Fei believes that there is still room for improvement in LingGuang, although in terms of final results.But this kind of AI scenario that takes the one-stop shop that really lowers the barriers to almost zero and solves a large number of needs is the trend that will drive the next step in the industry.

Meanwhile, LingGuang's multimodal delivery was recognized by many netizens. One netizen described it figuratively asOther Chatbot is still in the “txt stage”, while LingGuang has already done the graphic.

Aura supports the output of 3D, audio and video, charts, animations, maps and other full-modal information. Previously, we have already found in the real test that its multimodal answers make the dialog more vivid, with a clear focus, and the presentation form is very informative and aesthetic.

Some users believe that Aura has chosen a path of differentiation, and that the readability and ability to structure its answers is an important part of the user experience.Not all products need to become polarizingProductivity tools.

Other netizens have observed that non-techies around themselves can play with Aura this time around. This may have something to do with Aura's more user-friendly design in terms of UI.

Behind the real feedback and spontaneous creation fervor, a key question has also surfaced - what did Aura do right to quickly win users' favor in a crowded track?

Second, from the technical showmanship to the user experience, why Aura burst into flames?

Why do users think “Aura works well” and “Aura is different from other AIs”? It's not as simple as one more function or one more entrance, but Aura has stepped on the long-standing experience pain points in the industry and given the answer in a way that is closer to the public users.

In the past, the industry competed with model parameters, reasoning speed and other hard power, while “Aura” is the underlying ability to transform these into a visible, usable, immediately perceivable experience enhancement.

AI programming, Vibe Coding is one of the buzzwords in the AI industry in recent years, but when it comes to ordinary people, the palpable benefits are actually very limited: you can generate a bit of code, but you can't deploy it; you can write a demo, but it's hard to use it in real-life scenarios.There's always been a technological chasm between AI's capabilities and the creativity of its users.

“The ”flash application" has allowed many users to discover for the first time that this technological gap is actually not so difficult to overcome.Users do not need to write code, do not need to understand the front-end, do not need to solve the networking, tuning interfaces on their own these underlying complexity, and do not even need to turn on the computer, as long as the requirements are described, the aura will be able to turn it into a small application that can be used directly.

LingGuang lets users who don't know how to develop experience the fun of development for the first time, and also meets the long-neglected “fringe needs” - those personalized functions that belong to only a small circle and a small group of individuals, which can be solved in a few minutes in LingGuang. So even though these apps aren't perfect.They still give the user a sense of fulfillment in creating something with their own hands.

Another point that users clearly perceive as “good” is that Aura has taken a road less traveled in the way it presents information.

In the past year, with the explosion of inference models, Agents, AI-generated content has become significantly longer and more complex, which has admittedly brought about an increase in the depth and breadth of information, but also in a sense has evolved into a cognitive strain as well.

According to David Tsai, head of product at AuraThe presentation of information should not just be “cold text”, but should optimize the user's information access experience through a variety of displays.Aura's multimodal delivery is a response to this pain point.

What supports this experience is the “full code generation” path adopted by LingGuang. Although users see a simple and clear graphical interface, behind it is actually a large amount of code generated by multiple large models, and Aura presents complex content at a glance through code-driven diagrams, structured summaries, and information cards.

Users are not exaggerating when they describe it as “richly illustrated” and “like an article on a public website”. It really lowers the barriers to reading low enough.The answer to the big model is transformed from a mere information output to a complete “information product”.

Whether it is the release of creativity brought about by “flash applications” or the decline of cognitive burden brought about by multimodal presentation, they all point to the same trend: the competition of AI products is shifting from modeling capabilities to user experience. LingGuang's changes have made more non-technical users feel for the first time that AI is not just a tool, but can really improve the trivial aspects of life and is easy to manage. This is the key reason why it can win the recognition of a large number of real users in a short time.

Third, multimodal reconfiguration of information access, flash applications to promote AI universality

a miraculous column of light inMultimodal and user interaction experience approachThe exploration on is actually not unlike the direction of the current global head AI players.

For example, Google'sGeminiThe series of models has maintained a native multimodal orientation from the beginning. In this regard, Nobel Laureate Demis Hassabis, founder and CEO of DeepMind, explains that native multimodality is a key underpinning of Google's vision of a “universal digital assistant”. The latest update to Gemini 3 provides “dynamic view” capabilities, allowing Gemini to present information in cards, charts, and high-resolution images.

OpenAI has also recently tried to apply multimodality to information presentation. Its Pulse feature, launched at the end of September, attempts to let AI actively provide users with a tailored information card, as if it were carefully selected and produced by a human editor, combining text, charts, images and other information in a structured and intuitive way.

LingGuang, which has just been released, has been working on multimodal content for theUnderstanding and generationThe previously mentioned multimodal delivery is its initial result on the generation side.

LingGuang has also created a visual understanding function called “LingGuang Open Eyes”, and Cai Wei revealed that their positioning of “Aura Open Eyes” is as follows“Visual Native Portal”The purpose is to provide a natural, continuous mode of communication that is closer to human interaction.

The explosion of Aura Flash apps reflects the general trend of AI practicality, where ordinary users can quickly generate usable applets or tools according to their needs without programming or specialized skills, transforming ideas directly into reality. This concept is also echoed in Gemini 3's upgrades to its app development capabilities.

This trend towards more inclusive creation and lower barriers to entry also holds the potential to transform the relations of production in application development. In the past, development has been “centralized”: only a few programmers or teams of programmers have the expertise to design, implement, and deliver applications, leaving the average user with little access to off-the-shelf tools and no real participation in their creation.

the enhancement of the underlying modeling capabilities.It could make “development” as basic a skill as reading or writing.The software ecosystem is now being driven towards decentralization, personalization, and immediacy.

In previous communications with Wise Things and other media, Ant Group CTO He Zhengyu and Ant Lingguang head Cai Wei also revealed the future direction of Lingguang's development.

At the technical level, the flash application is still in its early form, but will become richer and more efficient over the next 6-18 months as the programming capabilities of the model and the capabilities of the AI mobilization tools continue to improve.

The core goal of Aura's technological evolution is to realize the quality of “more speed, better quality” to motivate users to create and share. At the same time, the team is experimenting with more interactive forms and modes, including SVG vector graphics, to present complex concepts.

In terms of product experience, in the future, LingGuang will continue to lower the threshold for the creation and use of flash apps, while exploring a complete app ecosystem and sharing capabilities. As the technology matures, flash apps are expected to naturally develop an app market and app ecosystem. Overall, LingGuang will be centered aroundEfficient messagingas the main axis, through theEnhanced Code Generation and Agent Capabilities, bringing about a qualitative leap.

Conclusion: Toward an AI-enabled future in AI-to-C experience innovation

Lingguang joins the battle of general-purpose AI assistants, adding a key piece to the domestic AI application map. Currently, domestic AI products have shown a clear division of the track: some are deeply engaged in content and entertainment, through lightweight interaction and fragmented experience to establish user stickiness; some are positioned as general assistants, covering a wide range of scenarios, providing intelligent Q&A and task execution; and Lingguang is based on the core of utility tools, oriented to the rigid needs of life and work, and emphasizes the product value of “able to do things, really useful! It emphasizes the product value of ”can do things, really useful".

With the continuous improvement of the model capability and increasing versatility, in the future, Aura is also expected to be deeply linked with the multifarious services of the Ant Alipay ecosystem, which will truly open up the complete closed loop from the user's “demand” to the “completion of the service”.

Technological breakthroughs are often rapid, yet it's the revolution of experience that truly changes the public's relationship with technology. Aura's latest exploration brings us one step closer to that transformation.

Source: Smart Stuff