Chinese AI Lab DeepSeek massively undercuts OpenAI on prices

• DeepSeek has rolled out AI models that are much cheaper than OpenAI’s offerings.
• AI models are mostly trained on public data, making differentiation challenging.
• DeepSeek’s open source models challenge OpenAI’s proprietary approach.

The cost of using AI models has come down as competition intensifies, making it increasingly easy to catch, match and even beat the best deals.

Having a leading AI model is not nearly as special as it was two years ago. These models are mostly trained on data that is publicly available on the Internet, so they are not that different from each other. It is becoming increasingly difficult to stand out in this crowd and charge premium prices.

In recent weeks, DeepSeek has taken this price dynamic to a new level. The Chinese AI lab rolled out models that are as good as or better than the best products from OpenAI, the pioneering creator of ChatGPT.

It is impressive. But what’s potentially more disturbing is how cheap DeepSeek’s models are.

Bernstein’s technology analysts examined DeepSeek’s offerings in recent days and found that the Chinese AI lab is massively undercutting OpenAI on price.

“DeepSeek’s pricing blows everything away from the competition, with the company pricing their models anywhere from 20-40 times cheaper than equivalent models from OpenAI,” the analysts wrote in a note to investors on Sunday.

The chart above shows the cost of “tokens,” which has become the raw material of generative AI. Chatbots and AI models break down words and other input into these tokens to make them easier to process and understand. A token is about three-quarters of a word.

Related stories

When AI companies handle prompts and other model inputs and outputs, they often charge users based on a cost-per-use basis. DeepSeek’s Reasoner model costs 55 cents for every 1 million tokens entered. Meanwhile, OpenAI’s o1 model charges $15 for the same number of tokens, according to Bernstein.

The Bernstein analysts also noted that DeepSeek’s models are open source, meaning they are available to anyone who wants to work with them for free. That’s a contrast to OpenAI, which keeps its top models proprietary and closed while charging relatively high prices for the products.

“It all raises some very interesting questions about the role and viability of proprietary versus open source efforts that are likely worth more work,” the Bernstein analysts wrote.

DeepSeek’s recently released flagship model is now the no. 1 on Apple’s free apps chart – just ahead of ChatGPT.