李开复创办的AI大模型独角兽公司零一万物近日宣布完成新一轮数亿美元融资，本轮融资吸引了包括某国际战投和东南亚财团在内的多家机构参与。这一轮融资不仅显示了市场对零一万物及其AI大模型技术的认可，也标志着公司全球化战略的进一步推进。
零一万物成立于2023年5月，由创新工场董事长、前微软全球副总裁李开复创立，核心团队成员来自谷歌、微软、IBM、百度等公司。公司自成立以来，便致力于打造领先的AI大模型技术，并在全球范围内寻求合作伙伴和投资者。
本轮融资完成后，零一万物估值跨过200亿元大关，成为行业内的领跑者之一。公司表示，将继续加大研发投入，加速产品迭代，并拓展国际市场。在C端市场，零一万物的AI办公工具PopAi已在海外市场取得初步成功，用户数近千万，并计划在今年内推出更多本土化产品。在B端市场，零一万物推出了首款闭源模型Yi-Large，获得了全球500强企业等高端客户的支持。
李开复表示，零一万物在技术成本和产品市场契合度（TC-PMF）方面取得了显著进展，这得益于对芯片计算水平、Infra计算效率、训练数据质量的优化。在5月21日的世界权威盲测LMSYS榜单中，Yi-Large综合排名中国模型第一，中文分榜世界第一。
尽管大模型行业面临集体降价的压力，零一万物坚持不盲目跟风，认为价格战会损害公司盈利能力和行业健康发展。公司强调，模型的定价应基于其技术价值，这需要模型厂商在人才建设上投入更多，特别是算法、Infra、工程三位一体的全球化人才。
未来，零一万物将继续深化全球化战略，不仅在本土市场推出AI产品，也在海外市场寻求更多合作机会。同时，公司将继续吸引全球顶尖AI人才加盟，以保持其在AI大模型领域的领先地位。
英语如下：
News Title: “Globalization Ambition of Zero to One: The AI Model Pioneer”
Keywords: Financing, Lee Kai-fu, Large Models
News Content:
Lee Kai-fu’s AI large-model unicorn company, Zero to One, recently announced the completion of a new round of funding amounting to tens of millions of dollars. This round of financing attracted participation from a variety of institutions, including an international strategic investor and a Southeast Asian consortium. This round of financing not only reflects the market’s recognition of Zero to One and its AI large-model technology but also marks a further advancement of the company’s globalization strategy.
Zero to One was established in May 2023 and was founded by Lee Kai-fu, the chairman of Innovation Works and a former Microsoft Global Vice President. The core team members come from companies such as Google, Microsoft, IBM, and Baidu. Since its inception, the company has been dedicated to developing leading AI large-model technology and seeking partners and investors worldwide.
Following the completion of this round of financing, Zero to One’s valuation has exceeded 20 billion yuan, making it one of the leaders in the industry. The company stated that it will continue to increase its research and development investment, accelerate product iteration, and expand into international markets. In the C-end market, Zero to One’s AI office tool PopAi has achieved initial success in overseas markets, with millions of users, and plans to launch more localized products this year. In the B-end market, Zero to One has launched its first proprietary model, Yi-Large, which has received support from high-end customers such as Fortune Global 500 companies.
Lee Kai-fu stated that Zero to One has made significant progress in terms of technical cost and product market fit (TC-PMF), which is attributed to the optimization of chip computing levels, Infra computing efficiency, and training data quality. On May 21, the Yi-Large model ranked first in China in the world-renowned blind test LMSYS list, and first in the world in the Chinese-language sub-list.
Despite the collective pressure of price cuts in the large-model industry, Zero to One insists on not blindly following the trend, believing that price wars will harm the company’s profitability and the healthy development of the industry. The company emphasizes that the pricing of models should be based on their technological value, which requires model manufacturers to invest more in talent construction, particularly in global talents specializing in algorithms, Infra, and engineering.
Looking ahead, Zero to One will continue to deepen its globalization strategy, not only launching AI products in the domestic market but also seeking more cooperation opportunities in overseas markets. At the same time, the company will continue to attract global top AI talents to join, to maintain its leading position in the AI large-model sector.
【来源】https://mp.weixin.qq.com/s/JwEHBtHs4HQZMCh5pg25Ng
Views: 2