Chinese AI startup DeepSeek has reached an annualized revenue run rate of $1 billion. This figure represents a more than twofold increase compared to the company's performance just a few months ago.
The growth in revenue follows a strategic decision to increase model prices by approximately 2.3 to 4.5 times. Sustained demand for the company's AI models has also played a significant role in this financial expansion.
DeepSeek is currently seeking to raise ¥50 billion, which is approximately $7.45 billion, at a valuation of ¥500 billion. The company intends to finalize this funding round by the end of October.
The startup has engaged CITIC Securities to explore a potential IPO on the Shanghai STAR Market. While preparations are underway, the specific timing, valuation, and scale of the offering remain undetermined.
Operational data indicates that DeepSeek prioritizes research and development, with over 70% of its computing capacity dedicated to training new models. The remaining 30% is allocated to inference for existing services.
The company recently launched the DeepSeek-V4.1-Flash model, designed to improve inference speed and throughput. This release aligns with the firm's focus on scaling its infrastructure for future model generations.
