AI news ·
Deepseek nears at least $12 billion funding round with CATL and Tencent backing
DeepSeek is raising at least $12 billion led by CATL and Tencent, targeting up to $15 billion after its V4-Flash model matched Western AI at lower cost.
DeepSeek is closing in on a funding round of at least $12 billion, with battery manufacturer CATL and internet giant Tencent leading the investment, Bloomberg reported. The Chinese AI startup originally sought $7.5 billion at a $75 billion valuation but may now target up to $15 billion as investor demand surges. The renewed interest follows the launch of its V4-Flash model, which matches or beats competing systems from Anthropic and OpenAI while operating at lower cost.
The V4-Flash catalyst
DeepSeek released preview versions of V4-Pro and V4-Flash in April. The V4-Flash model has drawn particular attention for outperforming Western rivals on key benchmarks without the price tag typically associated with frontier AI. Founder Liang Wenfeng has emphasized open-source development, offering freely available model weights - a strategy that contrasts with the proprietary approaches of OpenAI and Anthropic.
The company paused its fundraising earlier this year after viral criticism over its reliance on Nvidia chips. That round has since resumed, and in June DeepSeek closed its first external funding at $7.4 billion, valuing the company above $50 billion.
Infrastructure and chip independence
DeepSeek is building a data center powered by at least 160,000 AI chips from Huawei. The company is also developing its own inference chips to reduce dependence on both Nvidia and Huawei. These hardware moves signal a long-term bet on vertical integration, mirroring strategies seen at larger cloud providers but executed at a startup scale.
The company plans an initial public offering in 2027, according to people familiar with the matter. The combination of aggressive fundraising, infrastructure buildout, and chip development points toward a full-stack ambition that extends well beyond model research.
Why this matters for executives and strategy
DeepSeek's trajectory challenges assumptions about the cost structure of frontier AI. A startup offering competitive model performance at lower operational expense - while building its own silicon and data center capacity - reshapes procurement and partnership calculations for enterprises. For strategy leaders tracking the AI vendor landscape, the emergence of well-funded, vertically integrated Chinese labs adds a new dimension to build-versus-buy decisions and supply chain risk assessments. Professionals evaluating their team's readiness can explore AI Strategy Development Courses or AI for Executives Courses to strengthen their analytical frameworks around these shifts.