Bold Profitability Claims with Important Caveats
Chinese AI startup DeepSeek has made bold claims about the profitability of its AI models, though with significant caveats.
Theoretical Profit Margins and Revenue Projections
In a recent post on X, DeepSeek highlighted an impressive "cost profit margin" of 545% for its online services. However, this figure is based on "theoretical income," making it more of a projection than an actual financial snapshot.
DeepSeek’s Cost and Revenue Breakdown
DeepSeek elaborated on these calculations in a detailed GitHub post, where it outlined its strategy for enhancing efficiency and reducing latency. The company estimated that if all 24-hour usage of its V3 and R1 models were billed at R1 pricing, the daily revenue would total approximately $562,027. In contrast, the expense of leasing the required GPUs would amount to just $87,072.
The Reality Behind the Numbers
Despite these figures, DeepSeek acknowledged that its actual revenue is "substantially lower" due to factors such as discounted nighttime rates, reduced pricing for V3, and the fact that many services—including web and app access—remain free. While monetization of these services could improve margins, it would likely impact overall usage, making the projections somewhat speculative.
Broader Industry Implications
These statements come amid broader discussions surrounding AI's cost structures and long-term profitability. DeepSeek first gained widespread attention in January when it introduced a model that reportedly matched OpenAI’s o1 on key benchmarks—despite being developed at a fraction of the cost and under U.S. trade restrictions limiting access to advanced chips. This revelation caused tech stocks to decline and sparked industry-wide debates about AI investments.
Competitive Position in the Market
Beyond financial implications, DeepSeek's rise has also made waves in the consumer space. Its app briefly overtook OpenAI’s ChatGPT at the top of Apple’s App Store rankings. However, it has since dropped in overall rankings and currently sits at #6 in the productivity category, trailing behind ChatGPT, Grok, and Google Gemini.
Conclusion: Future Challenges and Opportunities
DeepSeek's claims underscore both the potential and challenges facing AI startups in a competitive, capital-intensive landscape. As the industry continues to evolve, profitability will depend not just on theoretical margins but on sustainable business models and strategic monetization efforts.
Stay updated with the latest insights—subscribe to our newsletter for real-time market analysis!
