On March 25th, Zhihu (02390) released its financial results for the fourth quarter and full year of 2025, delivering a landmark "profit report": the company's total annual revenue reached 2.75 billion RMB, with Q4 revenue at 644 million RMB. More critically, Zhihu achieved a full-year non-GAAP net profit of 37.9 million RMB, marking its first annual profit on a non-GAAP basis. This breakthrough not only signifies that Zhihu has maintained profitability for three consecutive quarters following its initial quarterly profit in Q4 2024, formally entering a new phase of "high-quality growth alongside comprehensive profitability," but also profoundly validates the long-term value of its synergistic strategy centered on "high-quality content × expert network × AI capabilities" in the AI era. At the critical juncture of its 15th anniversary coinciding with the AI explosion, Zhihu, through全面提升的经营质量 and the dual-drive of innovative businesses, has not only solidified the ecological moat of its professional Chinese internet community but also set a solid tone for initiating its next growth cycle.

The profit milestone represents a value reassessment and the realization of long-termism for "content assets" in the AI era. Zhihu's first annual non-GAAP profit in 2025 is seen as an inevitable value return for its deep cultivation of a "professional discussion" positioning and steadfast commitment to its community ecosystem. In the latter half of the internet era where traffic红利 are diminishing, Zhihu's financial report demonstrates that high-quality content and its expert network are not just moats for its business model, but core production factors capable of weathering economic cycles. Capital markets have often underestimated the long-term compound interest effect of professional content, tending to value content companies based on short-term user metrics or advertising growth rates. However, as we progress from the mobile internet to the AI era, the cost of generating information approaches zero, while "trustworthy answers" and "high-level professional judgment" become increasingly scarce. Zhihu's沉淀的, unique expert network and deep content repository within the Chinese internet constitute key infrastructure for calibrating AI hallucinations and providing high-value cognition. This scarcity is quantitatively verified in the latest report: Q4 paid reading business revenue was 334 million RMB, accounting for 51.8% of total revenue and remaining the largest revenue pillar. More notably, within the paid reading segment, the explosive growth of IP monetization validates the correctness of the "content assetization" path. The report shows that Q4 IP licensing revenue increased over fivefold year-over-year, and the number of authors earning版权 income grew sixfold annually. These figures clearly indicate the market is beginning to pay a premium for "professionalism," and Zhihu has successfully transformed previously undervalued content assets into tangible commercial cash flow. The leap from quarterly to annual profitability essentially represents the financial realization of Zhihu's long-termist strategy. Facing industry fluctuations, Zhihu did not sacrifice content quality for short-term traffic gains but continued to optimize the synergistic flywheel of "high-quality content × expert network × AI capabilities." A 24% quarter-over-quarter increase in marketing services revenue with improved structure in Q4, coupled with a stable monthly average of 12.2 million subscribing members, corroborates that a virtuous cycle of "good money driving out bad" has been established.

In the current landscape where generative AI is reshaping information distribution logic, Zhihu's profit breakthrough reveals a profound insight: AI does not dilute the value of human-generated content; instead, through "calibration effects" and "multiplier effects," it exponentially amplifies its scarcity. Zhihu is not merely a traffic arena vying for user attention but is committed to building a "human cognition base." Amid frequent algorithmic hallucinations, Zhihu's repository of cutting-edge discussions contributed by top practitioners such as Ma Huateng, Zhang Yitang, and the DeepSeek team has become "golden data" for training highly reliable AI systems. Research confirms that Zhihu leads community platforms in citation rates within professional fields, signifying that its content assets have溢出 from internal consumption to become infrastructure for the entire AI ecosystem, establishing an irreplaceable "data source" status. This "human + AI" collaborative paradigm is translating into concrete commercial growth momentum. Zhihu has proactively internalized AI capabilities as an operational engine rather than passively awaiting technological change. Its native search product "Zhihu Direct Answer" saw significant penetration growth in 2025, not only improving user efficiency in accessing high signal-to-noise ratio information but also bridging the commercial loop from "professional discussion" to "decision-making services." By upgrading its comprehensive search system, Zhihu successfully transformed vast amounts of professional content into structured assets understandable by machines and reusable across scenarios, drastically reducing the circulation cost of quality content. Zhihu's core competitiveness lies in its constructed "positive feedback system": using AI tools to empower creators and enhance production efficiency, while relying on the network effects of human experts to calibrate AI outputs, thereby building trust barriers in high-value scenarios like science education, consumer decision-making, and career development. This model proves that in an increasingly complex world, platforms that help people establish genuine connections and continuously upgrade their cognition will command valuation premiums that transcend cycles. Zhihu's profitability is not just a victory in operational efficiency but a milestone in the value reassessment of "trusted content" as a core production factor in the AI era.

Zhihu's growth logic has evolved from单纯的流量变现 to the deep synergy of "high-quality content × expert network × AI capabilities." The core lies in leveraging generative AI technology to reconstruct the value chain of content assets, unleashing substantial commercial增量. The standout Q4 IP licensing revenue, which surged over fivefold year-over-year, is a leading indicator of this strategic shift. Leveraging the vast high-quality short-story library of "Yan Yan Stories," Zhihu is establishing a difficult-to-replicate first-mover advantage and cost barrier in the AI-generated comic series (manju) track through the dual-drive of "IP licensing + self-produced incubation." Compared to traditional long-form video or live-action short dramas, Zhihu's short stories naturally possess characteristics like "short creation cycles, high plot density, and low adaptation costs," which create a perfect "chemical reaction" with the current technical features of AI video generation. Yan Yan Stories provides not only market-validated scripts but also ensures a high artistic floor for the manju through the text's inherent quality. This full-chain闭环 of "upstream content + midstream AI production + downstream multi-platform distribution"极大地提升了 the reuse efficiency and turnover speed of content assets. Specifically, the self-produced manju "Yao Zhuo Xue Hen" topped the Douyin charts within four days of release, surpassing 100 million views, while licensed works like "Lian Hua Lao" and "Rou Fei" dominated various platforms. By early 2026, copyrights for hundreds of works had already been sold. These hits are not accidental but the inevitable result of Zhihu leveraging AI technology to conduct a "dimensionality reduction strike" and "dimensionality elevation monetization" of its多年沉淀的文字 assets. Zhihu is evolving from a content community into an "AI-era content asset operator." By accelerating its布局 in AI manju, Zhihu has not only unlocked a third growth curve beyond advertising and membership but also achieved exponential amplification of content value: the same story can be transformed into text reading, audiobooks, short dramas, manju, and other formats, repeatedly capturing user attention and willingness to pay across different scenarios. This AI-driven "one fish, multiple dishes" model significantly reduces marginal production costs and notably improves gross margin and ROI. In the future, as Zhihu further deepens the application of AI in commercialization, its vast data advantages and IP reserves will transform into a sustained cash flow engine, propelling the company from "quarterly profitability" towards a new stage of "high-quality, sustainable growth."

In the grand context of AI reshaping information production relations, Zhihu's solid performance proves that only platforms capable of continuously accumulating high signal-to-noise ratio content,凝聚 real expert wisdom, and skillfully leveraging new technologies to amplify human cognitive value possess the resilience to穿越 economic cycles. From the deep cultivation of paid reading to the exponential爆发 of IP licensing revenue, from the重构 of search experience via "Zhihu Direct Answer" to the limitless extension of content boundaries through AI manju, Zhihu is undergoing a profound transformation from a "content community" to an "AI-era content asset operator." This is not only a rich reward for the long-term adherence to professionalism but also provides capital markets with a new valuation anchor—in an era of algorithm proliferation, "trust" is the most expensive luxury, and the "human element" is the most efficient calibrator. As the flywheel of "high-quality content × expert network × AI capabilities" accelerates, what Zhihu unleashes is not only commercial compound growth but also a more certain future for digital civilization in its pursuit of authenticity and depth. For investors, Zhihu's first profitable year may well mark the starting point of a sustained bull run in its value.