The Information：中國DeepSeek年化營收推算值達10億美元

（更正第4段：該公司是將透過募資籌措資金，並非透過首次公開募股（IPO））(Corrects paragraph 4 to say company to raise money via a fundraise, not an IPO)

路透9月24日電——科技媒體The Information週四報導，中國AI新創公司DeepSeek的年化營收推算值已達10億美元，較數月前增加逾一倍。該報導引述兩名直接知情人士說法。Sept 24 (Reuters) - Chinese AI startup DeepSeek's annualised revenue run rate has hit $1 billion, more than double from a few months ago, the Information reported on Thursday, citing two people with direct knowledge of the matter.

營收推算值顯示，如果一家公司維持目前表現，一年可創造多少營收。Run-rate revenue shows how much a company would make in a year if it continued performing at its current pace.

這項數據凸顯這家AI公司快速成長。DeepSeek去年因旗下模型挑戰美國對中國AI能力的既有假設，震撼科技業。該公司計畫很快上市，以支應擴張資金。The figure underscores the rapid growth of the AI company that shook the tech industry last year after its models challenged US assumptions about China's AI capabilities. The company is planning to go public soon to fund its expansion.

The Information報導，DeepSeek執行長梁文鋒在近期與投資人會面時分享了最新營收數字；該公司正推進第二輪募資，目標是在10月底前以5,000億元人民幣估值，籌集500億元人民幣（74.5億美元）。報導並稱，DeepSeek也正在為可能在上海證券交易所上市做準備。CEO Liang Wenfeng shared the latest revenue figure at a recent meeting with investors, as the company pushes ahead with a second funding round targeting 50 billion yuan ($7.45 billion) at a valuation of 500 billion yuan by the end of October, the Information reported. DeepSeek is also preparing for a potential listing on the Shanghai Stock Exchange, it added.

該新聞媒體表示，DeepSeek營收成長部分是由模型定價提高2.3倍至4.5倍所帶動。DeepSeek's revenue growth was partly driven by raising model pricing by 2.3 to 4.5 times, the news outlet said.

路透社無法立即查證這項報導。暫時無法取得DeepSeek回應。Reuters could not immediately verify the report. DeepSeek could not be immediately reached for comment.

路透社本月報導，DeepSeek已聘請中信證券，為可能在科創板首次公開募股做準備，不過時程、估值和募資規模仍未決定。DeepSeek had hired CITIC Securities to prepare for a potential STAR Market IPO, though the timing, valuation and fundraising size remain undecided, Reuters reported this month.

The Information並稱，梁文鋒告訴投資人，這家AI公司持續將大部分資源投入新模型開發，將逾70%的運算能力分配給訓練，保留不到30%用於推論，也就是運行既有模型。The AI firm continues to devote the bulk of its resources to developing new models, allocating more than 70% of its computing capacity to training while reserving less than 30% for inference, or running existing models, Liang told investors, the Information added.

本月稍早，該公司發布最新模型DeepSeek-V4.1-Flash，並表示該模型設計目標是提升能力、加快推論速度、提高吞吐量，並可擴展至更大型模型。Earlier this month, the company released its latest model, DeepSeek-V4.1-Flash, and said it is designed for greater capability, faster inference, higher throughput and scaling to larger models.

（1美元＝6.7131元人民幣）($1 = 6.7131 Chinese yuan renminbi)

（1美元＝6.7113元人民幣）($1 = 6.7113 Chinese yuan renminbi)

（班加羅爾Abu Sultan報導；Harikrishnan Nair編輯）(Reporting by Abu Sultan in Bengaluru; Editing by Harikrishnan Nair)