DeepSeek valued at $45 billion: The battle for AI talent behind its first funding round

Core conclusion

Chinese AI lab DeepSeek is raising its first external financing since its founding, valuing it at $45 billion. Founder Liang Wenfeng (who controls about 90% of the company's shares) aims to raise money to retain core talents through equity incentives - competitors are poaching DeepSeek researchers on a large scale.

Key Points

• Financing Round: First round of external financing since founding
• Valuation: ~$45 billion
• Financing motivation: Not for lack of money, but to allocate shares to employees to retain talents
• Background: DeepSeek shot to fame in early 2025 with large models that could be trained very cheaply

The Rise of DeepSeek: From “Price Butcher” to $45 Billion Valuation

DeepSeek was founded by Chinese hedge fund billionaire Liang Wenfeng. At the beginning of 2025, this laboratory entered the global AI horizon in a disruptive way-the open source large language model it launched achieved performance close to GPT-4 level at a computing cost and training budget far lower than that of its American counterparts (OpenAI, Anthropic). This achievement directly triggered the "DeepSeek moment" in the global AI industry and caused short-term fluctuations in the stock prices of many American AI companies.

Since then, DeepSeek has continued to iterate the model, and has successively released DeepSeek V2, DeepSeek V3 and the latest R2 (1.7 trillion parameter open source model). Each release has triggered a huge response in the open source community.

The war for talent: the real driver of financing

Contrary to common perception, the purpose of DeepSeek’s financing this time is not to “burn money to train a larger model.” According to the Financial Times, competitors are systematically poaching DeepSeek researchers. Liang Wenfeng chose financing in order to allow employees to hold shares in the company and retain the core AI research talents through equity binding.

This strategy is not uncommon in the AI ​​industry. In the past two years, from Google (Gemini team), Meta (Llama team) to OpenAI, talent flow has always been the core variable that determines the company's competitiveness. DeepSeek is famous for its low-cost training, and its researchers’ unique experience in “achieving higher performance with less computing power” is the scarce capability that giants crave.

Impact on AI development ecosystem

What this means for automated workflows

For developers and entrepreneurs who use AI tools to build automation and content production workflows, DeepSeek’s financing event sends several important signals:

1. AI model price competition will continue DeepSeek's continued presence ensures downward pressure on API pricing. Whether it is OpenAI’s continued price reduction of GPT-4o (a reduction of 60%) or DeepSeek’s low-price strategy, it is a direct benefit to producers of automated workflows.

2. The ecological value of the open source model is being recognized by the capital market The $45 billion valuation demonstrates the business value of the open source (or open weight) large model. This means that DeepSeek's models will continue to iterate, and users who use DeepSeek to build automation do not have to worry about the company behind it suddenly disappearing.

3. Attrition is the biggest risk for open source model companies The core asset of an open source model company is the team, not the code. If core researchers are poached, the iteration speed of subsequent models may be affected. This is a potential risk to consider when using DeepSeek models as production dependencies.

Tools related to DeepSeek

Tool brand names that naturally appear in the text will be automatically matched as tool_mentions:

Entries such as DeepSeek, OpenAI, ChatGPT, Anthropic, Claude, and Claude Code will trigger floating tool cards on the site to provide readers with more tool information.

Reference sources

• TechCrunch: DeepSeek could hit $45B valuation
• Financial Times (original reporting)

Internal link guidance

• Want to save money with DeepSeek? Watch the tutorial: Claude Code + DeepSeek V4 Building Tutorial: API Fees Dropped by 90% (15 Minutes)
• Existing case: Claude Code 48 hours to start a business: one person + US$29 monthly fee, monthly income in 3 months $9,000
• Matching tool: DeepSeek V4 vs Claude Code: 90% Cheaper, Same Quality