Newsletter Subscribe
Enter your email address below and subscribe to our newsletter
[forminator_form id="25163"]

TechNode+1WEEX+1ReutersDetails from a nearly four-hour closed-door investor meeting with DeepSeek founder Liang Wenfeng have circulated widely after Tencent Tech published a lightly edited transcript this week. In it, Liang laid out a strategy centered entirely on achieving artificial general intelligence, rejecting the path of building a consumer super-app or competing for traffic with ByteDance or other Chinese tech giants.
Liang told investors that DeepSeek will not pursue 3D generation, video generation, or world models, describing those as useful applications but not central to the development of intelligence. "Restraint is a strategy," he said, according to the transcript. "Sometimes you give up some things to gain more of others."Hello China Tech+2
He outlined a development roadmap moving from chain-of-thought reasoning to AI agents, continual learning, and eventually embodied intelligence. Coding agents are the company's most important near-term investment, Liang said, because they could improve both DeepSeek's products and its own research process. He also committed to keeping DeepSeek's strongest models open source, arguing that closed-source approaches offer no meaningful advantage.TechNode+2
The disclosures come after DeepSeek completed its first-ever external funding round, raising approximately 51 billion yuan ($7 billion), according to Pandaily. Tencent contributed around 10 billion yuan, with battery maker CATL also participating as a major investor. The round valued the company at roughly 350 billion yuan.Reuters+3
DeepSeek is now preparing for a potential mainland IPO on Shanghai's STAR Market. Reuters reported on July 15 that the company has set an internal target to complete an IPO filing by the end of 2026, with a listing potentially following in 2027. The company is also in discussions for a fresh funding round at a valuation of about 500 billion yuan ($74 billion) ahead of the public offering.Reuters+1
Liang was candid about the constraints facing Chinese AI development. "The biggest gap between us and the US is in resources," he said, describing chip procurement difficulties and high prices. He expressed optimism about China's domestic chip ecosystem, predicting that within one year, real-world deployment would demonstrate its viability — and describing Huawei's 950 super-node as capable of replacing Nvidia hardware in performance and price.Hello China Tech
He framed DeepSeek's famed efficiency not as a philosophical choice but as an adaptation to scarcity, telling investors: "We believe in scaling. Bigger is always better. What stops us from scaling is compute, not desire."Hello China Tech