DeepSeek is back at the fundraising table. The Hangzhou-based AI startup has resumed its second external funding round, targeting roughly $8 billion at a valuation approaching 500 billion yuan, or about $74 billion. Hong Kong-based Monolith Management is in active talks to join the round.
For context on how fast this company has moved: DeepSeek was founded in 2023, employs around 200 people, and is now being valued at more than most publicly traded banks.
What happened, and why the pause
DeepSeek closed its first external funding round in June 2026, pulling in over $7.4 billion at a post-money valuation above $50 billion. That round attracted some of China’s biggest corporate names, including Tencent and JD.com, alongside Monolith Management.
The second round hit a wall in July 2026 after comments made by founder Liang Wenfeng to investors were leaked. The fallout was enough to halt the process entirely, at least temporarily.
Now the round appears to be back on track. Monolith Management, which has previously backed rival AI startup Moonshot AI, is in discussions to participate.
Liang Wenfeng himself made the largest single contribution to the first round, committing roughly $3 billion personally.
Why DeepSeek keeps attracting capital
DeepSeek’s core product is open-source large language models and multimodal AI technologies.
The company’s valuation jumped from above $50 billion after the first round to approaching $74 billion now, a roughly 48% increase in the span of a few months.
DeepSeek first grabbed global attention earlier in 2025 when its models demonstrated performance competitive with Western frontier models at a fraction of the reported training cost. That moment rattled US tech stocks and sparked genuine debate about whether the capital-intensive approach to AI development was the only viable path.