Following a brief and dramatic suspension, DeepSeek has officially restarted its massive second external funding round, seeking to raise capital at a valuation that would cement its position as one of the most valuable artificial intelligence startups in the world. According to a report by Bloomberg, the Chinese AI developer is currently seeking close to $8 billion from investors, targeting a pre-money valuation approaching 500 billion yuan, or approximately $74 billion. The resumption of the fundraising effort indicates that the company has successfully navigated recent internal turbulence and is moving aggressively to secure the financial resources required to maintain its disruptive pricing strategy and train next-generation models.
The targeted $74 billion valuation represents a staggering 48 percent increase from the valuation achieved just months ago. DeepSeek closed its first-ever external funding round in June 2026, raising roughly $7.4 billion at a valuation of approximately $59 billion. The rapid escalation in the company’s perceived market value underscores the intense demand among institutional investors for exposure to China’s leading foundation model developers, particularly those that have demonstrated the ability to compete with top-tier American systems while operating under severe semiconductor export controls. Among the firms currently in talks to participate in the new round is Monolith Management, signaling strong interest from prominent investment vehicles.
(Related: DeepSeek Seeks $7.4 Billion in First External Funding Round at Up to $59 Billion Valuation)
Recovering from the July Pause
The resumption of the funding round marks a critical stabilization for DeepSeek after a highly publicized disruption late last month. The company had initially paused this second $8 billion round on July 26, following a series of viral social media posts by its founder, Liang Wenfeng. The posts, which circulated widely across Chinese tech circles, reportedly caused friction with potential investors and prompted the company’s leadership to temporarily halt negotiations to manage the fallout. The fact that the round has been restarted so quickly, and at the same massive valuation target, suggests that the underlying confidence in DeepSeek’s technological roadmap and commercial viability remains intact.
DeepSeek’s ability to command such a premium valuation is rooted in its unique market positioning. The company has built a reputation for delivering near-frontier AI capabilities at rock-bottom prices, severely undercutting both domestic rivals and Silicon Valley incumbents like OpenAI and Anthropic. Its models have consistently performed exceptionally well on independent benchmarks for coding, mathematics, and reasoning, proving that Chinese developers can achieve world-class results through algorithmic efficiency and architectural innovations like Mixture-of-Experts (MoE), even without unfettered access to the latest Nvidia hardware. This combination of high performance and low inference costs has driven massive developer adoption and forced a broader price war across the global AI ecosystem.
The Capital Intensity of the AI Race
The sheer scale of DeepSeek’s fundraising ambitions, seeking over $15 billion across two rounds in a single summer, highlights the brutal capital intensity of the current phase of the AI revolution. Training frontier models requires tens of thousands of advanced graphics processing units (GPUs), massive data center infrastructure, and immense energy resources. For Chinese companies, this challenge is compounded by the need to source alternative hardware or pay steep premiums on the gray market to acquire restricted American chips. To remain competitive, companies must secure multi-billion-dollar war chests simply to cover the cost of compute.
By targeting a $74 billion valuation, DeepSeek aims to permanently differentiate itself from the crowded field of mid-tier Chinese AI startups and establish itself as a peer to global tech giants. The capital raised in this round will be essential not only for purchasing hardware but also for funding aggressive expansion into new domains, such as multimodal vision-language models and embodied intelligence for robotics. This strategic diversification is already underway, as evidenced by DeepSeek’s recent $20.8 million strategic investment in the humanoid robot manufacturer Unitree, a partnership designed to pair DeepSeek’s cognitive models with Unitree’s physical machines, a deal announced alongside Unitree’s IPO pricing on the Shanghai STAR Market.
(Related: DeepSeek Pauses $1.5B Second Funding Round After Viral Liang Wenfeng Posts)
Reshaping the Global Startup Landscape
If DeepSeek successfully closes this $8 billion round at the targeted valuation, it will reshape the hierarchy of the global AI startup ecosystem. A $74 billion valuation would place the Chinese developer in the same rarefied air as OpenAI, which was valued at over $80 billion earlier this year, and well ahead of other prominent Western labs like Anthropic and Cohere. It would also serve as a powerful validation of China’s domestic AI industry, proving that despite aggressive US efforts to throttle the country’s technological advancement, Chinese firms can still attract the massive capital necessary to compete at the frontier.
For the broader market, DeepSeek’s fundraising success ensures that the brutal price war currently defining the AI sector will likely continue. Armed with billions in fresh capital, the company can afford to sustain its aggressively low API pricing for the foreseeable future, subsidizing user acquisition while it scales its infrastructure. This dynamic places immense pressure on competitors worldwide to either match DeepSeek’s efficiency or risk losing market share to the heavily capitalized Chinese disruptor. As the round moves toward a close, the implications of DeepSeek’s growing financial muscle will be felt from Beijing to Silicon Valley.