# Pricing Related Articles

HTX News Center provides the latest articles and in-depth analysis on "Pricing", covering market trends, project updates, tech developments, and regulatory policies in the crypto industry.

Understanding the New Economic Model of Tokenization

Understanding the New Token Economics Model The commercialization of AI applications is evolving from selling software and subscriptions to selling token call capacity. Tokens, the fundamental unit of information processing for large language models (LLMs), have become the basis for API billing and consumption. With call volumes exploding, tokens themselves are now being traded—procured, routed, split, and resold—forming a new intermediary market. This layer connects upstream LLM providers with downstream developers and enterprises, acting as a global wholesale-to-retail liquidity network. The rise of this business is fueled by a massive surge in China's daily token call volume—growing over a thousandfold from 100 billion in early 2024 to over 140 trillion by March 2026—and significant improvements in domestic LLM capabilities, which are now competitive globally. The core value of token distribution platforms extends beyond simple arbitrage. Key functions include aggregating multiple models (like GPT, Claude, and domestic models such as Kimi and DeepSeek) under a unified API, lowering network and payment barriers, and providing enterprise services like model selection, prompt engineering, and system integration. Profit models are diversifying: (1) resale margins; (2) technical premiums from proprietary inference acceleration (e.g., reducing costs to 1/10 of the industry standard); and (3) enterprise value-added services. High-consumption scenarios like marketing, short-form video, gaming, and e-commerce are primary drivers. Investment opportunities are seen in both companies with strong model capabilities (e.g., Alibaba, Tencent, MiniMax) and those with high-consumption client scenarios (e.g., marketing agencies with overseas reach). However, risks are significant: low entry barriers leading to intense competition, capital requirements and bad debt risks from advance payments, and dependency on policy changes from upstream LLM providers who control API pricing and access.

marsbit05/19 02:54

Understanding the New Economic Model of Tokenization

marsbit05/19 02:54

China's AI Circle Has Just Established a Pecking Order, and Capital Is Already Changing the Rules Again

The article describes how the valuation logic for major Chinese AI model companies has undergone three dramatic shifts between 2022 and 2026, driven by capital's changing priorities. The first phase (around 2022) was **technology-driven valuation**, where funding was based on model performance and benchmark scores. This logic was disrupted when DeepSeek's R1 model demonstrated that comparable capabilities could be achieved at a fraction of the cost, challenging the notion of technical superiority as an unassailable moat. The second phase shifted to **IPO window-driven valuation**. Following favorable listing conditions in Hong Kong, capital flowed to companies like Zhipu and MiniMax with the clearest path to a public listing. However, this focus on liquidity over fundamentals became apparent as their Annual Recurring Revenue (ARR) lagged far behind international peers like Anthropic. The third and current phase is **national strategy-driven valuation**. This shift was marked by the state-backed "Big Fund" leading a major investment in DeepSeek, signaling that leading domestic AI models are now viewed as strategic national assets comparable to semiconductor manufacturing. This new logic, combined with soaring US valuation benchmarks (e.g., OpenAI at $850B), propelled the combined valuation of China's top AI firms ("The Four Dragons"/"Five Strong") past 1 trillion RMB. The article presents a "pricing leap model": each shift is triggered by a key event that invalidates the old logic, leading to rapid capital reallocation under a new narrative before its flaws (particularly the gap in fundamental ARR metrics) become evident. It concludes that the next major test for these valuations will be a return to scrutinizing core business fundamentals, specifically ARR growth, suggesting a fourth pricing shift is imminent.

marsbit05/18 10:42

China's AI Circle Has Just Established a Pecking Order, and Capital Is Already Changing the Rules Again

marsbit05/18 10:42

Introducing a 'Paid Subscription' in the Chinese Market, What's Doubao Thinking?

Chinese AI assistant "Doubao" (from ByteDance) has announced it will launch a paid subscription service alongside its free version, with plans priced at 68, 200, and 500 yuan per month. This move follows its achievement of over 345 million monthly active users and 1.8 billion daily interactions. The paid tiers aim to serve professional users with advanced features for complex tasks like PPT generation and data analysis, while basic functions remain free. The timing is strategic: user growth from free services is plateauing, and the market is now more receptive to paying for high-value AI tools. ByteDance leverages its technical edge in model efficiency and cost control to support this shift. However, significant challenges remain. The Chinese market is characterized by low long-term subscription loyalty, with users often paying only for immediate needs. Doubao's premium features face competition from free alternatives offered by rivals. Furthermore, the core business model of AI subscriptions struggles with scalability—more paying users mean higher compute costs, potentially creating a cycle where revenue fails to cover expenses. Intense price competition from rivals could also force difficult choices between maintaining premium pricing or engaging in a race to the bottom. In summary, while Doubao's massive user base ensures short-term subscription uptake, its long-term success depends on creating uniquely valuable, "sticky" services within ByteDance's ecosystem and solving the fundamental industry dilemmas of low renewal rates and unsustainable cost structures. The outcome will serve as a critical test case for the viability of premium C-end AI subscriptions in China.

marsbit05/14 02:50

Introducing a 'Paid Subscription' in the Chinese Market, What's Doubao Thinking?

marsbit05/14 02:50

Why Pricing Social Interactions is Doomed to Fail?

Titled "Why Putting a Price on Social Interaction Is Doomed to Fail," this article critiques attempts to monetize social networks directly through SocialFi models, arguing their inevitable failure stems from a fundamental misunderstanding of media dynamics. Using Marshall McLuhan's theory of "hot" and "cold" media, the author posits that social networks are inherently "cold" media. Their value isn't contained in individual posts but is co-created through user participation, interpretation, and fragmented, ongoing interaction (e.g., replies, shares). This ambiguity and need for user involvement are core to their function. The article asserts that SocialFi projects like Friend.tech failed because introducing real-time, tradable financial pricing (a definitive "hot" signal) into this "cold" environment doesn't add a layer—it replaces the medium's essence. The unambiguous price signal overshadows and nullifies the nuanced, participatory social signal. Users become traders, not participants, and when speculative profits vanish, the underlying social ecosystem—never genuinely cultivated—collapses entirely. This principle extends beyond crypto. The author argues platforms like Twitter have gradually "heated up" through metrics (likes, retweets counts, algorithmically defined value), shifting users from participants to performers and eroding organic engagement. The solution isn't to abandon capital but to manage its entry point. Successful models like Substack, Patreon, or Bandcamp allow capital to "condense" at specific, isolated nodes (e.g., subscriptions, one-time payments) without permeating and "heating" every social interaction. They preserve the core "cold," participatory medium while enabling monetization at designated boundaries. The NFT boom and bust serves as a stark parallel: the ancient "cold" medium of collecting (valued for story, community, gradual accumulation) was rapidly destroyed by platforms that introduced real-time floor prices, rarity scores, and trading dashboards, transforming collectors into speculators and vaporizing cultural value when prices fell. The core lesson: "Liquidity equals heat." Injecting high liquidity and definitive pricing into a "cold" participatory medium doesn't optimize it; it fundamentally alters and destroys its value-creating mechanism. The future lies not in pricing every social gesture but in finding precise, non-invasive points for capital to condense without overheating the entire ecosystem.

marsbit05/11 13:11

Why Pricing Social Interactions is Doomed to Fail?

marsbit05/11 13:11

DeepSeek No Longer Wants to Focus Only on Large Models

DeepSeek, a leading Chinese AI company, has released its new model series DeepSeek-V4, featuring two versions: the high-performance V4-Pro with 1.6 trillion parameters and the cost-efficient V4-Flash. Both support 1 million token context windows and use Mixture-of-Experts (MoE) architecture to improve efficiency. The company continues its strategy of offering competitive pricing, with input tokens priced as low as ¥0.2 per million tokens. A key revelation is DeepSeek’s explicit link between future price reductions and the mass availability of Huawei’s Ascend 950 AI chips in the second half of the year. This signals a strategic shift from relying solely on algorithmic and engineering optimizations to integrating domestic computing power into its core cost structure. DeepSeek has adapted its inference system to run efficiently on both NVIDIA GPUs and Huawei NPUs, potentially challenging NVIDIA's CUDA ecosystem dominance. Concurrently, DeepSeek is reportedly seeking significant external investment, with a pre-money valuation of around ¥300 billion. This move highlights growing pressures in scaling compute infrastructure, retaining top talent—amid recent departures of key researchers—and accelerating commercialization efforts. The company has also updated its consumer app with tiered model access, indicating a stronger product focus. The V4 release underscores that China's AI competition is evolving beyond pure model capability into a broader contest involving compute supply chains, engineering systems, financing, and talent strategy.

marsbit04/25 01:45

DeepSeek No Longer Wants to Focus Only on Large Models

marsbit04/25 01:45

OpenAI Goes Left, DeepSeek Goes Right

On April 24, 2026, DeepSeek released V4, a Chinese large language model offering a free "million-token context window," enabling it to process vast amounts of data like entire books or years of corporate documents in one go. In contrast, OpenAI’s GPT-5.5, released around the same time, is more powerful but significantly more expensive, charging up to $180 per million output tokens. DeepSeek’s strategy represents a shift from a pure AI research firm to a heavy-infrastructure player, building data centers in Inner Mongolia’s Ulanqab to bypass U.S. chip export restrictions. This move, supported by Huawei’s Ascend chips and China’s cheap green electricity, highlights a fundamental divergence in AI development models: U.S. firms focus on high-cost, high-margin services, while Chinese players like DeepSeek prioritize accessibility and affordability. Facing intense talent poaching from tech giants, DeepSeek is seeking a $44 billion valuation funding round to retain researchers and scale infrastructure. Meanwhile, Chinese manufacturers are compressing AI models to run on smartphones, making AI accessible offline and across the Global South. Through open-source models and localized solutions, Chinese AI is empowering non-English speakers and low-income users, driving a form of "digital equality." While Silicon Valley builds walled gardens, DeepSeek and others are turning AI into a public utility—like tap water—flowing freely to those previously left behind.

marsbit04/24 07:33

OpenAI Goes Left, DeepSeek Goes Right

marsbit04/24 07:33

AI "Transfer Station" Earning Millions Monthly? Five Questions Uncover the Truth of Token Arbitrage

The article "AI 'Transfer Station' Earns Millions Monthly? Five Questions Uncover the Truth of Token Arbitrage" explores the emerging business of API token transfer stations, which profit from global AI service price disparities and access barriers. These intermediaries purchase low-cost tokens from overseas AI providers (e.g., OpenAI, Claude) through grey-market methods—such as exploiting enterprise credits, bulk accounts, or subscription benefits—and resell them to Chinese users at a markup. Key drivers include the high cost of using top AI models (e.g., Claude Code costs ~$5 per million tokens), the performance gap between domestic and foreign models, and mismatches between subscription and API pricing. However, the practice carries significant risks: upstream token sources may be unstable or illegal; user data passing through intermediaries can be harvested or injected with hidden prompts; and models might be downgraded without disclosure. The market is evolving, with some operators now exporting cheaper Chinese models (e.g., Qwen3.5 at ~$0.11 per million tokens) to overseas users, leveraging price gaps. Yet, sustainability is low due to compliance crackdowns, instability, and reputational risks. Users are advised to employ detection methods (e.g., prompt adherence tests) and avoid sensitive data usage. The authors caution that while transfer stations offer short-term arbitrage, they lack long-term reliability and security compared to official APIs.

marsbit04/24 00:26

AI "Transfer Station" Earning Millions Monthly? Five Questions Uncover the Truth of Token Arbitrage

marsbit04/24 00:26

活动图片