What happened
Tang Hanlin, chairman of DataBao, said competition in the token market is currently focused on the cost and supply capacity of general-purpose tokens. But as AI applications deepen, customer demand will shift from simply having tokens available to caring about actual usage results.
He described the token industry as likely going through two phases. The initial phase centers on general-purpose tokens, with customers focused on availability, price, and stability. The second phase, he said, will see customers demand tokens tailored to specific scenarios, giving an example of medical AI: a general model can answer some medical questions, but a model further developed with specialized medical knowledge and data could win users even if its token price is higher.
DataBao has recently upgraded from a data-element operator to a data-native AI token service provider. The company says it now segments and classifies demand, matches different models to different needs, and uses DataBao, ZhisuanBao, and CiyuanBao to cover data, computing power, and tokens for the same group of B-end clients.
Why it matters
The remarks point to a shift in how token value is assessed: from generic supply metrics to domain-specific application outcomes. If the market moves in that direction, providers with proprietary, professionally curated data may gain pricing power even when their token costs are not the lowest.
The strategy also suggests that companies in the data and AI chain are broadening their positioning from raw data services to integrated offerings that include data, computing resources, and token access, aiming to serve enterprise customers across multiple layers of the AI stack.
Key facts
Token competition currently centers on the cost and supply capacity of general-purpose tokens.
Customer needs will shift from having tokens available to focusing on actual use effects, especially in professional scenarios.
DataBao has upgraded from a data-element operator to a data-native AI token service provider.
DataBao covers the data, computing, and token links through DataBao, ZhisuanBao, and CiyuanBao, targeting the same batch of B-end customers.
What to watch next
Whether DataBao can strengthen its position in specialist fields such as medical AI, where domain data and model refinement become competitive differentiators.
How the company's evolution from data APIs to token APIs and then computing-power APIs plays out in customer adoption and value delivery.