DeepSeek AI

What is DeepSeek AI?

DeepSeek is a Chinese based startup AI company created just over a year ago by hedge fund manager Liang Wenfeng. His hedge fund, named High-Flyer, used AI chips to build algorithms to identify “patterns that could affect stock prices,” wrote the Financial Times. He’s recently been compared to Sam Altman of OpenAI. Liang’s ambitious developments have already sent shockwaves through the world of artificial intelligence.

DeepSeek-R1 Launch

Launched on January 20th, 2025, DeepSeek’s latest model, DeepSeek-R1, has taken the AI landscape by storm. The model competes directly with top-tier offerings from OpenAI and Claude, but at a fraction of the price. While OpenAI’s premium ChatGPT services cost between $20 and $200, DeepSeek’s model charges just $0.14 per million tokens. This dramatically undercuts competitors for similar performance and capability.

The Fallout

This launch was very much unexpected and has upset the American tech industry by showing they may not be the leaders of AI innovation for much longer. It is also an open-source platform meaning it allows developers worldwide to access the code and innovate on its technology.

By the morning of Monday the 27th, DeepSeek had rocketed to the #1 app in the Apple App Store and the US stock market had seen over $2 trillion in losses with companies such as Nvidia seeing losses up to 17%, their largest ever loss in a single day.

DeepSeek attracted global attention with their claim that their V3 model (DeepSeek-R1 predecessor) required less than $6 million worth of computing power to create and was reportedly trained using 10,000 Nvidia H800 Chips. It showed the global market that you don’t need to be a massive player to compete in the AI space and begs the question, if true, as to why so much has been flooded into the industry in the US with a similar outcome.

DeepSeek AI Services

DeepSeek AI offers a range of individual services, each designed to harness the power of AI in their unique ways. Below are the core services currently offered by DeepSeek:

• DeepSeek Chat
• DeepSeek Coder
• DeepSeek Math
• DeepSeek LLM

How Can These Be Leveraged for Businesses?

DeepSeek Chat

This model is ideal for enhancing customer support and engagement. DeepSeek Chat can automate emails, run your whole CMS website, provide 24/7 assistance, analyse documents and much more. Usually, to make sure you can automate these services, you will need a third party service such as make.com.

DeepSeek Coder

The coder model is pretty self explanatory. It can assist software development teams by generating code snippets, identifying bugs, and offering suggestions to optimise coding processes. It can work with all programming languages and ultimately can reduce the workload of developers.

DeepSeek Math

For industries that require complex mathematical computations, such as finance or engineering, DeepSeek Math can perform advanced calculations and provide accurate solutions.

DeepSeek LLM

LLM stands for Large Language Model. It is an advanced language model comprising 67 billion parameters. It was designed to understand and generate human-like text, enabling various applications, such as content creation, natural language understanding, and decision-making assistance.

Final Thoughts

Since AI models burst onto the scene, the USA has been the driving force of innovation. The launch of DeepSeek AI disrupts this, which causes a major shift in the AI landscape. Personally, I think an extra competitor in the space is great for the industry and reduces the opportunity for monopolies. It also reduces complacency in the industry, forcing quicker innovations to occur.

The security side of a Chinese Artificial Intelligence sparks concerns, however, the leaders in US, OpenAI have also come under fire for similar controversy with Sam Altman changing it from a non-profit company and making the code private. DeepSeek-R1 being released as an open-source model was a smart play from Liang and increases trust with consumers. By making the code accessible, DeepSeek positions itself as a transparent and community-driven leader in the AI space. This currently creates a shift towards DeepSeek models, however, the open-source nature allows these competing companies to build based off their code.

In conclusion, DeepSeek’s latest R1 model marks a pivotal moment in the AI industry. It will be intriguing to see how other AI companies evolve and develop better models in the near future. Personally I hope it sets an industry standard with its open-source transparency.