Chinese artificial intelligence startup DeepSeek has begun laying the groundwork for a mainland initial public offering while simultaneously exploring a new private funding round that could push its valuation to approximately $71 billion, according to reports from the Financial Times and Bloomberg. The Hangzhou-based company has engaged accounting and banking advisers to prepare financial reports, with a possible IPO filing near the end of 2026 or early 2027.
The fundraising discussions come just weeks after DeepSeek closed its first external financing round, which raised about $7 billion and resulted in a post-money valuation of around $52 billion. The new round targets a pre-money valuation of 480 billion yuan ($71 billion) and aims to secure at least 10 billion yuan in fresh capital, though the final amount could increase if more investors participate.
DeepSeek has gained international recognition for its V3 and R1 large language models, which demonstrated strong performance and lower reported training costs, challenging assumptions that Chinese AI firms cannot compete with U.S. leaders like OpenAI and Anthropic. The company is now expanding into agentic AI, which enables systems to handle complex tasks with greater autonomy, and is developing its own AI chips to reduce reliance on third-party hardware.
The startup, founded in 2023 and backed by entities including Tencent, CATL, and China’s National Artificial Intelligence Industry Investment Fund, will focus on advanced research over near-term commercial returns, according to founder Liang Wenfeng. Liang’s net worth has surged to an estimated $36 billion following the funding activity. DeepSeek’s IPO ambitions align with a broader trend of AI companies moving toward public markets, with both OpenAI and Anthropic reportedly filing confidential IPO documents in 2026.
