A groundbreaking financial product has officially launched in Guangzhou, marking the first batch of its kind nationwide. This innovative lending solution, designed for AI companies lacking traditional physical assets, provides access to credit based on computing power consumption metrics, with individual borrowers able to secure up to 30 million yuan in credit lines.
The product represents a fundamental shift in how banks evaluate loan applications. Traditional lending practices typically emphasize physical collateral such as residential properties, factory buildings, machinery, and the personal asset accumulation of business owners. In contrast, this new credit offering pivots to consider alternative indicators including a company's historical token settlement data, token supply and consumption values, and the valuation of their computing power contracts.
This approach allows financial institutions to assess a company's genuine operational performance through their computing power usage patterns, gradually transforming the lending logic from static asset evaluation toward dynamic data monitoring. A senior executive at the Guangzhou branch of the Bank of China highlighted that computing power has become a key reference point for determining credit limits, enabling banks to better understand the real production and operating conditions of their clients.
Three distinct loan products
The product line is structured into three sub-categories: supply loans, application loans, and service loans, each designed to address different operational needs within the AI ecosystem. The maximum credit available to any single borrower reaches 30 million yuan with a three-year term. Since its introduction, the Bank of China's Haizhu District branch has already extended credit to six companies, cumulatively totaling 28 million yuan.
Industry analysts note that this kind of financial innovation significantly restructures traditional credit risk control logic for the banking sector. It enables banks to better reach asset-light AI technology startups and other such client groups while simultaneously pushing financial institutions to refine their risk assessment models, data verification processes, and post-loan management procedures. This development effectively challenges banks to evolve their approaches to accommodate the changing nature of business assets in the digital era.
AI startups find relief from cash flow pressures
Several innovative companies across different sectors have already benefited from this new lending mechanism. One AI short drama production company, facing substantial token consumption costs, has secured a 5 million yuan credit line with a three-year term. The company's operational costs illustrate the financial intensity of their business: producing standard quality short dramas requires approximately 1,500 yuan in token costs per minute, while top-tier quality productions escalate to 4,000 yuan per minute. With order volumes steadily increasing, the urgency of their financing needs has become apparent.
Another enterprise specializing in AI-powered e-commerce livestreaming has experienced rapidly growing monthly token usage. The company faced significant cash flow gaps because computing power procurement payments must be made in advance, while project payment cycles extend five to six months. Through this new credit product, they have obtained a three-year credit line of 3 million yuan, effectively alleviating their temporary funding pressures.
The significance of this computing power token loan extends beyond solving immediate financing problems. Experts in the internet research field point out that it introduces a new paradigm for evaluating business value in the AI era. As the nature of corporate assets continues to evolve, the metrics used to assess companies must adapt accordingly. If this model can be continuously improved and refined, it could play a highly constructive role in supporting technological innovation across the broader economy.