DeepSeek Advances 70 Billion Yuan Financing; Liang Wenfeng (梁文锋) Pledges to Prioritize Open‑Source AI over Quick Commercialization

DeepSeek (深寻科技) is reportedly moving forward with a massive financing plan worth about 70 billion yuan (roughly US$9.7 billion), even as its leadership signals a clear strategic choice: build open‑source AI models first, monetize later. It has been reported that Liang Wenfeng (梁文锋), who leads the company, has publicly pledged to continue developing and releasing open‑source foundational models rather than chasing short‑term commercialization gains. Why bet on the commons when money is on the table?

Funding and scale

It has been reported that the proposed 70 billion yuan package would support model training, data infrastructure and talent recruitment, positioning DeepSeek among the better‑funded players in China’s fast‑moving AI startup ecosystem. Such a war chest would allow large‑scale compute commitments and expanded R&D cadence. Reportedly, some of the capital could come from both private investors and strategic partners within China’s tech and industrial network.

Strategy amid geopolitical pressure

Liang’s open‑source emphasis comes against a backdrop of tighter US export controls and growing pressure on Chinese firms to accelerate domestic AI self‑reliance. Open sourcing can broaden community contributions, speed model iteration and reduce dependence on proprietary foreign toolchains — all attractive advantages as global tech policy fragments. It has been reported that DeepSeek sees open models as a longer‑term moat rather than an immediate revenue engine.

If true, DeepSeek’s stance raises questions for incumbent giants and rivals — Baidu (百度), Alibaba (阿里巴巴) and others are already racing on proprietary and hybrid models. Will open‑source foundational models reshape who captures value in China’s AI market? For now, Liang’s pledge signals a bet on ecosystem building over fast cashing out.