The Dispute over "Byte Time": The Scale and Boundaries of Super Giants

This week, a debate on "Who does the Chinese internet actually belong to" has accidentally uncovered the deepest structural changes taking place in China's technology industry.

Renowned scholar Qin Shuo published an article arguing that the Chinese internet has entered the "ByteDance era"; while Li Liang, Vice President of Douyin Group, released a response overnight, pointing out that there are common-sense errors in the statistical caliber used, and denied the monopoly characterization by noting that ByteDance's total duration share among the Top 50 Apps does not reach the one-half threshold specified in the Anti-Monopoly Law. Li Liang also specifically mentioned in his response that if Tencent's equity-invested Apps are included, the total usage duration would far exceed that of others. Following the same logic, the Chinese internet could be described as in the "Tencent era".

This Rashomon-style dispute over statistical caliber and business boundaries appears to be a remote dialogue between public relations practitioners and scholars, but the huge resonance it has triggered precisely reveals an unavoidable reality: regardless of whether the "ByteDance era" characterization is rigorous, ByteDance has grown into a super giant whose scale, wealth creation and technological depth are sufficient to define the standards of the entire industry.

Discussing the present and future of this company can no longer be limited to the daily active users of a single product or the statistical standard of a certain data. We need to break it down from deeper commercial logic, technological evolution and industrial ecology: how a super giant completes its scale leap and reconstructs the value coordinate of the entire industry in the gap between the transition from mobile internet to the AI era.

Rewriting Pricing Rules

When discussing the scale of a giant, the most direct measurement is the pricing of the capital market and the founder's wealth value.

In the private equity market in 2026, ByteDance's valuation ranges from 500 billion to 600 billion US dollars; Tencent's market value in the secondary market is about 504 billion US dollars in the same period, but it should be noted that Tencent's market value once exceeded 5 trillion Hong Kong dollars in March 2026 before a significant correction, and ByteDance's "approaching" part benefits from the low ebb of its competitor. If ByteDance lands on the public market with a valuation of around 600 billion US dollars, it will directly surpass Tencent and become the internet company with the highest market value in mainland China — but the primary market valuation is not equal to the listing market value, and Hong Kong stock IPOs often have discounts, so 600 billion US dollars is an "anchor" rather than a "result".

Along with the soaring valuation is the founder's rise to the top of the wealth list. The 2026 Hurun Rich List shows that 43-year-old Zhang Yiming's wealth increased by 200 billion yuan, making him the richest man in China for the second time with 670 billion yuan, and refreshing the wealth record of China's richest man. Hurun clearly pointed out that this growth directly benefits from the higher company valuation driven by the progress of the AI business. As for 55-year-old Pony Ma, his wealth shrank by 26% to 346 billion yuan, ranking fourth.

This is not just a change in the ranking of two wealthy individuals, but a fundamental shift in the logic of wealth creation in the Chinese internet industry.

In the past decade, Tencent's wealth myth was built on an unparalleled "social network effect". The acquaintance relationship chain built by WeChat and QQ has extremely high migration cost and network lock-in effect. This effect has brought Tencent rich game agency revenue and advertising profits, but it also means that the traffic dividend based on social relationships has reached a physical ceiling.

ByteDance's wealth leap represents another more explosive model: algorithm-driven attention distribution and full-stack technological dividend. From Toutiao to Douyin, then to Hongguo Short Drama and Doubao, ByteDance does not rely on users' social relationship chains, but on accurate calculation of users' dopamine and curiosity.

Financial data confirms this explosiveness. In the first half of 2026, ByteDance's revenue was about 120 billion US dollars, a year-on-year increase of about 30%, and its revenue scale has caught up with Meta's 117.1 billion US dollars in the same period; its net profit was about 20 billion US dollars, a year-on-year decrease of single-digit percentage, and overseas revenue accounted for more than 30%. Even though the net profit margin has declined due to the crazy investment in AI infrastructure, its cash flow and financing capacity are still sufficient to support it to maintain extremely high-intensity continuous investment in computing power, talents and infrastructure.

Full-Stack Closed Loop

The core driving force for ByteDance to catch up with and even surpass traditional giants in terms of valuation and wealth comes from its full-stack layout and extremely strong engineering implementation capabilities demonstrated in the AI era. This set of layouts does not stay at the conceptual level, and it leaves verifiable figures in every link.

At the model layer, the call scale is expanding at an exponential rate. As of June 2026, the daily average Tokens call volume of the Doubao large model has exceeded 180 trillion, which is more than 1500 times higher than the initial 120 billion scale when it was released in May 2024, and more than 10 times higher in the past year. In the public cloud MaaS market, Volcano Engine's Token share reaches 49.5% — for every two Tokens consumed by Chinese enterprises, one is provided by Volcano Engine. The number of enterprise customers with annual cumulative Token consumption exceeding 1 trillion has increased from 100 in December 2025 to 200.

At the application layer, the density of national-level products forms the traffic base of the closed loop. As of September 2026, the monthly active users of the Doubao App reached 419 million, a year-on-year increase of 143.6%; there are caliber differences in different sources of daily active user scale: the daily active users monitored by QuestMobile at the end of September are about 181 million, and other reports claim that the daily active users have exceeded 200 million. Doubao firmly ranks first in the domestic AI native application market. The daily active users of Hongguo Short Drama reached 168 million in July 2026, a year-on-year increase of 107%, which has exceeded the sum of the daily active users of iQiyi, Tencent Video, Youku and Mango TV (about 155 million); its 368 million monthly active users correspond to the data of June 2026. Together with Douyin, TikTok and Fanqie Novel, ByteDance holds super entrances with over 100 million daily active users in three dimensions: content consumption, social entertainment and AI assistant.

At the commercialization and tool layer, the growth rate of the revenue engine is also considerable. ByteDance raised its MaaS revenue target for 2026 to 15 billion yuan, which is 10 times its actual revenue in 2025 (about 1.5 billion yuan); among them, the Seedance 2.0 video generation model alone can contribute more than 1 billion yuan in revenue per month. Supporting all this is ByteDance's AI capital expenditure of more than 200 billion yuan in 2026, and a syndicated loan of about 29.6 billion US dollars.

Putting these figures together, the outline of the closed loop becomes truly clear: 180 trillion daily average Token calls drive the model capabilities of Doubao and Hongguo; 419 million monthly active users and 168 million daily active users generate massive real interactive data, feeding back model iteration; Seedance's 1 billion monthly revenue and Volcano Engine's 15 billion MaaS target convert technological capabilities into cash flow to support the next round of computing power investment. Models, traffic, data and capital form a high-speed cycle link in this system, and the scale effect of each link reduces the marginal cost for the next link.

This is the most intuitive manifestation of "the strong get stronger" in the AI era. While other large model companies are still anxious about the high computing power cost and the difficult-to-close-loop commercialization, ByteDance has completed the self-sustaining and continuous iteration of large model technology through extremely rich implementation scenarios. Every content distribution, every ad click, and every API call provides massive data feedback for its algorithm model.

Tension of Concentration

When an organization has the largest user duration share, the most complete AI product matrix, the most abundant cash flow and the most efficient monetization machine at the same time, resource concentration becomes a natural evolutionary result.

From a positive perspective, this concentration brings extremely high industrial efficiency. ByteDance uses recommendation algorithms to reconstruct the cost structure of content distribution, uses AI to significantly lower the production threshold of categories such as short dramas, and uses a unified traffic and commercialization system to make the turnover speed of creativity and production reach a level that traditional industries cannot match.

But resource concentration also raises tensions worthy of continuous observation.

When more and more attention, creative production capacity and commercial budgets converge to the same system, the diversity of the ecosystem, the bargaining space of small and medium participants, and the boundary of platform rules on creative freedom will all become long-term public issues. Li Liang's denial of the monopoly characterization in his response by stating that ByteDance's total 41.8% duration share among the Top 50 Apps does not reach one half is an effective defense in terms of legal and statistical caliber; but from the perspective of industrial ecology, the influence radius beyond the share figures — the traction on the upstream and downstream, the shaping power on pricing and distribution rules — also deserves to be included in the discussion.

In the practice of internet anti-monopoly, "one half" is the legal quantitative threshold, but it is not the only dimension to measure market influence. When a company builds a cross-domain network through equity participation, incubation, technology export and traffic distribution, its actual impact on market competition is often more complex than the share figure of a single App. ByteDance is not the first giant to face this kind of scrutiny. Tencent, Alibaba, Meta and Google have all experienced similar inquiries in the process of growing into infrastructure-level companies. This is a structural situation that any enterprise reaching a super scale must face.

Folding of Algorithms

If we examine ByteDance's expansion in the AI era at a deeper level, we will find that it is not only reconstructing business, but also reshaping the production relations of content.

Taking Hongguo Short Drama as an example, the core reason why it can surpass the traditional film and television industry in a short time is that it has opened up the complete chain of "web novel IP + AI production + algorithm distribution + advertising monetization". Fanqie Novel provides data-verified popular IPs, the Doubao large model acts as an efficient screenwriter, AI video tools further reduce production costs, and finally the works are accurately pushed to the audience through the algorithms of Hongguo and Douyin.

This model is extremely efficient, but it also brings a hidden kind of folding. Driven by algorithms, content creation is increasingly inclined to "data-verified pleasure points". Whatever theme has a high retention rate will be mass-produced; whatever reversal can stimulate dopamine will be stacked crazily. Algorithms have no bias, they just faithfully implement the principle of maximizing efficiency. But in this process, high-quality niche content that requires slow-paced foreshadowing, in-depth thinking, or does not conform to the public's pleasure points, is often marginalized due to insufficient ROI.

This is not a unique phenomenon of ByteDance, but a common challenge faced by the entire algorithm recommendation era. When platform rules and traffic distribution become an invisible baton, there is a natural structural tension between the independence of creators, the diversity of content, and the commercial efficiency pursued by the platform.

Boundaries of Giants

Whether the statement of "ByteDance era" is accurate may not be important. What is important is that ByteDance has proven itself to be one of the peaks of efficiency and scale in this era with its solid valuation, wealth, revenue and full-stack AI capabilities.

Zhang Yiming becoming the richest man, and ByteDance's valuation approaching or even possibly surpassing Tencent, is not something that needs to be praised or criticized in itself — they are the result of the combined effect of market, technology and organizational capabilities, and a presentation of the choice of this era.

In the long history of business, we often have an instinctive vigilance against behemoths. But objectively, the change of every era is inevitably accompanied by the birth of super giants. Microsoft in the PC era, Tencent in the mobile internet era, and ByteDance today are all representatives of the most advanced productive forces of their respective eras.

The rise of ByteDance represents the commercial height a Chinese enterprise can reach under the extreme talent density and extreme engineering efficiency. Relying on its precise insight into user needs, crazy investment in AI technology, and extremely fierce internal competition mechanism, it has broken out a new territory. It represents the real extremely competitive level of Chinese technology enterprises in overseas expansion and AI application layer.

But at the same time, when a company grows to be able to influence the information acquisition habits of hundreds of millions of people and determine the survival status of a large number of creators, its business decisions are no longer only related to financial statements, but have extremely extensive social attributes.

The balance between scale and boundary is the ultimate proposition that all super giants cannot avoid. How to find the dynamic balance between the pursuit of extreme commercial efficiency and the maintenance of ecological diversity, between the impulse of technological expansion and the constraint of business ethics, not only tests the wisdom of the management, but also determines whether this company can continuously obtain the trust of the public and the market in a longer time dimension.

The answer is not in any article, but in the evolution of market rules, the collision of industry rules, and the long-term echo of every business choice.

This article is from the WeChat official account "Tech Business", author: Tech Business, published with authorization from 36Kr.