The recent launch of DeepSeek, a Chinese AI startup, has shaken up the tech world. With the launch of their chatbot DeepSeek-R1 in December, they have presented a model that should excel existing models in both price and quality.
What makes this achievement remarkable is that with a relatively modest budget of less than $6 million, DeepSeek has developed an AI model that rivals the advanced models of established players such as OpenAI. This raises questions about the huge investments currently being made in the AI industry and suggests that innovation does not necessarily depend on huge budgets.
The rise of DeepSeek has not only generated admiration, but also concerns. Microsoft and OpenAI have made allegations about possible data theft by DeepSeek, further fuelling tensions between the US and Chinese tech sectors.
What does this mean for Europe?
These developments highlight the need for Europe to rethink its position in the AI world. The AI game has not yet been played. Europe must seize its moment now to play a significant role in this technological revolution.
But, there are also internal challenges. Excessive regulation within the European Union can hamper innovation, putting Europe at risk of falling behind in the global AI revolution.
Europe, however, has leading research centres, a strong science base and an ethical approach to AI development. But if it wants to compete with the US and China, it needs a strategic focus on collaboration, investment and scalability.
Flexibility in AI models
At Ittes, we recognise these challenges while seeing the opportunities they present. Our bespoke AI platform, IttesAI, is designed to provide businesses with flexible and scalable AI solutions to suit their specific needs. Innovation remains central for us, which is why we are currently exploring the possibility of adding DeepSeek to our AI platform. DeepSeek’s efficiency and performance make it an interesting option as an additional language model within IttesAI, especially for applications where cost efficiency is an important factor.
The right balance between reliability and cost efficiency
While we take into account that DeepSeek, while cheap and powerful, is less reliable than established models such as those from OpenAI or Anthropic. Therefore, the addition of DeepSeek within IttesAI may be especially useful for non-critical data and less sensitive applications, where cost efficiency outweighs absolute accuracy and reliability.
In fact, with IttesAI, you have the freedom to decide which language model to deploy, without having to take out multiple subscriptions. So you can flexibly switch between different models such as ChatGPT, Gemini, Claude and DeepSeek, depending on the specific application.
“By taking a hybrid approach, we can strike the right balance between affordability and quality.”
This means that companies can choose which model is most suitable for each scenario. For applications where customer data needs to be read in and reliability is paramount, a model like GPT-4 or Claude, for example, can be deployed. For less critical applications, where cost efficiency and speed are more important, it is easy to switch to a cheaper model such as DeepSeek. This flexibility makes it easier to optimally align AI with different business processes, without additional costs or complex licensing structures. By taking a hybrid approach, we can strike the right balance between affordability and quality.
The AI race is not over yet
The rise of DeepSeek serves as a wake-up call for Europe. It is a call to invest not only in technology, but also in collaboration, regulation and ethics. Through joint efforts, we can create an ecosystem in which European AI solutions flourish and contribute to a sustainable and inclusive digital future.
At Ittes, we remain committed to innovation and growth, with a strong focus on delivering value to our customers and strengthening Europe’s position in the global technology landscape.
