Liang Wenfeng, the visionary founder of DeepSeek, previously announced that the company boasts ample financial resources and is now intensifying its efforts towards capitalization. Its latest funding round has witnessed an overwhelming response, with subscriptions surpassing 100 billion yuan. This influx is predominantly from industrial capital, aimed at refining the shareholder composition and forging an industrial collaboration network. Factoring in the initial funding round, DeepSeek will have amassed over 150 billion yuan in capital within a mere six-month span. The pre-investment valuation for this round stands at approximately 500 billion yuan, with pricing remaining grounded despite the prevailing market enthusiasm. DeepSeek is gearing up for a listing on the Science and Technology Innovation Board in late August 2026, plans to submit its prospectus by the end of the year, and aims to go public in 2027. It has also recruited Yan Wentao, a former partner at Hillhouse Capital's venture capital division, to serve as its Chief Financial Officer. At present, leading global AI firms are ramping up their capitalization strategies, intensifying the competition for resources in the large model sector. DeepSeek remains committed to an open-source philosophy, recently open-sourcing its internal inference framework, DeepInfer, and production-grade codec component, DeepSeek-recipe. This move is designed to slash industry-wide computational power and energy consumption costs, while fostering the widespread adoption of large models. Liang Wenfeng is of the opinion that AGI cannot be the exclusive domain of a single entity and that exercising restraint and cultivating an open ecosystem are more conducive to achieving the ultimate objective. In terms of monetization, DeepSeek maintains a measured approach, charging solely through API channels at competitive rates. This strategy enables the recovery of hardware costs within ten months, without the pursuit of profit maximization. Currently, its annualized revenue has skyrocketed from under $500 million to $1 billion within a three-month period, catapulting it into the top tier of the domestic large model industry. Simultaneously, the company is incrementally introducing search-based paid interfaces to diversify its business offerings, with the relevant data also being utilized to enhance model training. At present, DeepSeek's primary focus is to propel recursive self-iteration in AGI research and invest in related fields of embodied intelligence, laying the groundwork for the future integration of intelligence into the physical realm.