DeepSeek's Annual Revenue Surpasses $1 Billion, Plans to Raise ¥50 Billion According to The Information, DeepSeek's annual revenue has exceeded $1 billion, more than doubling from less than $500 million just a few months ago. CEO Liang Wenfeng recently stated at an investor meeting that after increasing model API prices by 2.3 to 4.5 times, customer retention has not been significantly affected. DeepSeek aims to complete its second round of financing by the end of October, targeting ¥50 billion in fundraising, with a pre-money valuation of approximately ¥500 billion. Currently, the company's revenue is almost entirely derived from API services, as its free chatbot has yet to generate commercial income. In the first seven months of this year, DeepSeek's API business achieved a gross margin of 82.9%, higher than Anthropic and OpenAI during the same period. Over 70% of the company's computing power is allocated to training new models, with inference computing accounting for less than 30%. The company is also testing the use of gaming GPUs to run smaller models to alleviate pressure on inference resources.