IIn the world of artificial intelligence, the rise of DeepSeek is offering European companies a significant opportunity to level the playing field. Hemanth Mandapati, the CEO of the German startup Novo AI, was among the first to shift from OpenAI’s ChatGPT to the Chinese AI model, DeepSeek, just two weeks ago. Speaking at the GoWest conference in Gothenburg, Sweden, Mandapati explained how easy it was to migrate.
“If you’ve already built your app with OpenAI, migrating to other models is simple… it only takes us minutes,” Mandapati said in an interview.
Follow THE FUTURE on LinkedIn, Facebook, Instagram, X and Telegram
DeepSeek’s entry into the AI landscape is having a significant impact, particularly on pricing models. Interviews with startup leaders and investors reveal that the company’s affordable pricing structure is forcing competitors to reconsider their pricing and improve their models. According to Mandapati, DeepSeek’s pricing is five times lower than what competitors offer.
“DeepSeek offered pricing that was five times cheaper than competitors,” Mandapati explained. “I’m saving a lot of money, and users won’t notice any difference.”
European startups have long faced challenges in keeping pace with their American counterparts, primarily due to easier access to funding and resources. However, with DeepSeek’s cost-effective technology, European companies now have a chance to close the gap.
“This is a huge step toward democratizing AI and leveling the playing field with major tech giants,” said Seena Rejal, CEO of Netmind.AI, a UK-based company and one of DeepSeek’s early users.
Research from Bernstein analysts shows that DeepSeek’s pricing is 20 to 40 times lower than OpenAI’s. For example, OpenAI charges $2.50 for every $1 million in input tokens, while DeepSeek charges just 0.014 dollars for the same amount.
Despite the promising advantages, there are regulatory concerns. DeepSeek is under investigation in several European countries to determine whether it has copied data from OpenAI or if it is censoring responses to avoid negative portrayals of China.
A Shift In The AI Market
In 2024, the U.S. saw nearly $100 billion in venture capital investments in AI companies, while Europe only managed $15.8 billion, according to PitchBook data. Meanwhile, U.S. President Donald Trump recently unveiled Stargate, a $500 billion joint venture between OpenAI, SoftBank, and Oracle.
In Europe, investments in AI remain modest. However, some companies, like France’s Mistral, are managing to compete with the major players such as OpenAI, Meta, and Google. DeepSeek caught attention after it was revealed that the cost of training its DeepSeek-V3 model was less than $6 million using NVIDIA H800 computing power, making it one of the most affordable AI models to date.
“This shows that bigger isn’t always better,” said Fabrizio del Maffeo, CEO of Axelera AI. “As AI models become more accessible, costs fall, and barriers to innovation decrease, accelerating industry development.”
While some analysts question whether DeepSeek’s training costs are as low as reported, there’s no doubt that they are significantly cheaper than their U.S. counterparts. Ulrik R-T, CEO of Empatik AI, a Danish startup, sees DeepSeek as an opportunity for companies without large budgets.
“It proves we don’t need enormous budgets to realize our vision,” R-T said.
The Price War Begins
The shift in pricing has already triggered changes in the industry. Recently, Microsoft announced it would offer its OpenAI-powered logical reasoning model for free to Copilot users, a departure from its usual $20 per month subscription fee.
“AI prices are falling, so future solutions are likely to focus on more transparent, open-source models—even if they come from China,” said Joachim Schelde of Scale Capital.
However, larger corporations like Nokia and SAP are more cautious about these developments. According to Alexandru Voica, head of the corporate department at Synthesia, a UK-based company valued at $2.1 billion, price is just one factor.
“Other considerations include security certifications and software ecosystems that allow companies to integrate AI solutions into their platforms,” Voica added.
