DeepSeek Revenue Nears $500 Million as Chinese AI Startup Eyes IPO

DeepSeek Revenue Nears $500 Million as Chinese AI Startup Eyes IPO

Chinese artificial intelligence startup DeepSeek has reportedly built an annualized revenue base approaching $500 million, strengthening its plans to raise billions of dollars and pursue a public listing.

DeepSeek’s annualized revenue recently reached between $400 million and $500 million, driven largely by sales of cloud-based access to its models through application programming interfaces, The Information reported, citing three people familiar with the matter.

The 3-year-old AI company is seeking to raise 50 billion yuan, or about $7.4 billion, at a valuation of 500 billion yuan, or roughly $74 billion, according to the report. That would value DeepSeek at approximately 148 times its annualized revenue.

DeepSeek has maintained gross margins of 70% to 80% on access to its V4 flagship model despite charging less than leading U.S. competitors, The Information reported. Infrastructure improvements have allowed the company to process more queries with fewer chips, reducing the cost of running its models.

The economics support DeepSeek’s broader pitch that powerful AI models can be developed and operated with less computing infrastructure.

PYMNTS reported in February 2025 that DeepSeek charged $2.19 to generate 1 million output tokens through its API, compared with $60 for OpenAI’s o1 model at the time, making DeepSeek nearly 27 times cheaper.

PYMNTS also reported that businesses were adopting DeepSeek’s models because of their lower costs despite continuing national security and data privacy concerns. Enterprises could reduce exposure by self-hosting the models, while use of DeepSeek’s consumer application created greater data-sharing risks.

The Information reported that DeepSeek is particularly interested in attracting Middle Eastern investors.

The company has hired investment banks to prepare for a potential initial public listing on Shanghai’s STAR Market. It aims to file an application this year and complete the IPO next year, according to The Information.

The move represents a shift for founder Liang Wenfeng, who previously resisted external capital and emphasized frontier research over commercialization, according to the Wall Street Journal. It also shows how the need to finance compute is pushing even open-source AI developers toward venture and public markets.