DeepSeek raises $7.35 billion in first funding round, surpassing Alibaba! Seeking capital with the "fewest strings attached"

Chinese AI startup DeepSeek has launched its first round of external financing since its inception, aiming to raise up to $7.35 billion (50 billion RMB). The post-money valuation is expected to exceed $51.5 billion (350 billion RMB), marking the largest single financing round in the history of China's AI industry. Negotiations with Alibaba fell through due to demands for ecosystem integration, and Tencent’s proposal to acquire a 20% stake was declined. Founder Liang Wenfeng plans to personally contribute up to 20 billion RMB, accounting for approximately 40% of this round. (Previous coverage: Jensen Huang: Neither Blackwell nor Rubin should be accessible to China; NVIDIA remains firm despite zero market share in China) (Background: Anthropic spends $200 billion on Google Cloud over five years; two AI startups consume half of the orders from the four major cloud giants) - DeepSeek's first external financing round reaches up to $7.35 billion (50 billion RMB); founder Liang Wenfeng personally contributes 20 billion RMB for a 40% stake. - Negotiations with Alibaba failed due to demands for ecosystem binding; China's Big Fund is expected to lead the round with the most lenient terms. - V4.1 is scheduled for release in June with native MCP support, but core researchers have defected to Xiaomi and ByteDance. According to The Information, this Chinese AI startup, incubated by the quantitative fund High-Flyer Quant, has initiated its first round of external financing, aiming to raise up to 50 billion RMB (approx. $7.35 billion), with a projected post-money valuation exceeding 350 billion RMB (approx. $51.5 billion). If completed, this will be the largest single financing round in the history of China's AI industry. DeepSeek, which has avoided external financing for years, has officially opened its doors. However, even with the doors open, founder Liang Wenfeng is seeking "money with the fewest strings attached." The largest investor in this round is not an external institution, but founder Liang Wenfeng himself. He plans to personally invest up to 20 billion RMB, subscribing to approximately 40% of the equity. The China Integrated Circuit Industry Investment Fund (the AI-specific fund under the third phase of the Big Fund) is expected to become the second-largest investor, offering relatively lenient terms. Both Alibaba and Tencent actively negotiated but were turned away. Alibaba hoped to strengthen its AI ecosystem closed-loop through the investment, deeply integrating business units like Tongyi Qianwen into product lines such as Taobao and Amap. DeepSeek insisted on technical independence and refused ecosystem binding, leading to a breakdown in negotiations. Tencent proposed acquiring up to 20% of the shares, but was declined due to the excessive stake requested. Liang Wenfeng has long rejected external equity financing. Previously, DeepSeek was entirely supported by High-Flyer Quant's own capital, with Liang holding approximately 84% of the shares directly and indirectly. Although he has opened the door for the first time, his bottom line is to accept only offers with the "fewest additional conditions." The Big Fund's terms are the most lenient and best align with DeepSeek's roadmap. The pace of commercialization driven by this financing is accelerating. DeepSeek plans to release the V4.1 update in June, focusing on strengthening enterprise-side tools, native support for MCP (Model Context Protocol), and adding image and audio processing capabilities (generation remains text-only). This is the first time DeepSeek has explicitly targeted the enterprise market with tool-level products. However, talent attrition is more urgent than the roadmap. V3 core contributor Luo Fuli jumped to Xiaomi last November to lead the MiMo large model team; the MiMo-V2.5 released under her leadership this April has already outperformed DeepSeek-V4-Flash in several benchmarks. Another core researcher, Guo Daya, joined ByteDance's Seed team this April as the head of the Agent division (L8 rank), with rumors of a salary "near 100 million RMB" (ByteDance denied the specific figure). Reports indicate that ByteDance has poached nearly 70 people from DeepSeek over the past year. Competitors are offering salaries more than double those at DeepSeek. A major use of the new capital is to increase employee compensation and issue options, establishing market-based pricing for equity through external financing to provide liquidity for paper wealth. As recently as February, DeepSeek publicly stated it was "fully funded by High-Flyer, with no immediate financing plans." Three months later, soaring compute costs and a brutal talent war forced a change in stance. Comparing with US model makers, the cloud compute contracts of Anthropic and OpenAI combined have reached nearly $2 trillion. While