Deepseek claims the 545 % “theoretical” profit margins
Chinese Deepseek AI recently announced that artificial intelligence models may be very profitable – with some star.
In a post on X, Deepseek boasted that its online services have a “cost profit margin” by 545 %. However, this margin is calculated based on “theoretical income”.
These numbers discussed in more detail at the end of the GitHub tallest publication that determines its approach to achieving “higher productivity and decrease in cumin”. The company wrote that when you look at the use of V3 and R1 models within 24 hours, if The bills of this use were released using R1 prices, Deepseek will have already 562.027 dollars of daily revenue.
Meanwhile, the cost of renting graphics processing units (graphics processing units) was only $ 87,072.
The company admitted that its actual revenues are “largely less” for a variety of reasons, such as night discounts, low -pricing V3, and the fact that “only a sub -group of services”, with access to the web and the application remains free.
Of course, if the application and the website are not free, and if other discounts are not available, then the use is assumed to be much lower. So, these accounts seem very speculative – more gesture towards possible future profit margins than a real snapshot of Deepseek at the present time.
But the company shares these numbers amid broader discussions about the cost of artificial intelligence and possible profitability. Dibsic jumped to the spotlight in January, with a new model that is supposed to match Openai’s O1 on certain criteria, despite developing them at a much lower cost, and in the face of American commercial restrictions that prevent Chinese companies from reaching the strongest chips. Technology shares fell and analysts asked questions about the spending of artificial intelligence.
Deepseek technology is not only the presence of Wall Street. Its application has briefly operated Chatgpt from Openai at the top of the Apple App Store – although it has then decreased from the general classification and is currently ranked 6th in productivity, behind ChatGPT, Grok and Google Gemini.