What is DeepSeek AI? Why the Whole World is Talking About It?

There is news trending all over social media, new sites, and on television that you may know about. Yes, it’s DeepSeek AI, a China-based AI lab whose headquarters is in Hangzhou, Zhejiang, China. This lab recently launched its new AI model, DeepSeek R1, which is more effective than ChatGPT. After its launch in the market for Apple iOS and Android, the tech sector has lost almost $94 billion. Furthermore, it also wiped $600 billion in Nvidia's market value. But if a chatbot is affecting the market in such a way, then how can it be useful? Well, it is because this AI chatbot is totally free to use and an open-source large language model. 

What is DeepSeek? 

DeepSeek is a China-based artificial intelligence firm founded by Liang Wenfeng in May 2023. The firm recently launched its AI model, DeepSeek R1. This model has surpassed OpenAI’s GPT-4o and o1 with its free-to-use feature. Furthermore, its recently launched models are made on Nvidia’s lower-performing H800 chips. The chatbot gained popularity after releasing on Apple and Google’s app store. It also became the top-rated free application available on Apple's App Store, which was earlier the ChatGPT. 

History of DeepSeek AI 

In 2016, Liang Wenfeng, the CEO of the company, started trading during the 2008 financial crisis while studying at Zhejiang University. He created the Hedge Fund by using AI to make money in the stock market. Liang used AI only for trading and bought tons of powerful Nvidia chips (A100s) before the US government restricted them for China. Furthermore, in May 2023, he founded DeepSeek, an AI chatbot. He made sure that users got open-source and free-to-use large language models. The codes can be used, modified, and viewed by anyone. 

Moreover, then he released a very powerful but cheap AI model, DeepSeek-V2, in May 2024 at the lowest price, which forced other big companies like ByteDance, Tencent, Baidu, and Alibaba to lower their prices too. Despite earning less, the company was gaining popularity because the competitors were losing money, too. They hired college graduates and freshers to run their company. On January 10, 2025, the company released its latest AI model, DeepSeek-R1, which dominated the market. 

How Much Better is DeepSeek Compared to Other LLM Models?

DeepSeek AI vs. ChatGPT

DeepSeek AI and ChatGPT are both AI chatbots designed to help humans who work on natural language processing. However, ChatGPT 4 is expensive, and DeepSeek is free to use. Furthermore, ChatGPT generates user-friendly language, whereas R1 uses technical terms while giving answers. GPT is also the go-to for everyday users and interacts like humans. DeepSeek is useful for completing the technical tasks. 

DeepSeek AI vs. BERT 

BERT (Bidirectional Encoder Representations from Transformers) is an AI chatbot useful for helping people with automated language understanding. Furthermore, DeepSeek AI is also useful for handling technical queries and solving problems quickly. 

DeepSeek AI vs. Gemini

Compared to Gemini, DeepSeek is much better. It is because it generates answers faster than any other AI chatbot. Gemini has an advanced version, which costs $22.45 (Rs 1950) per month. On the other hand, DeepSeek AI offers all the premium features for free. It is an open-source large language model, which means anyone can access and use the code. 

Is it Safe to Use DeepSeek?

Despite being free to use, it is not important that it is safe to use. Just after the release of DeepSeek V3, it was hit with a large-scale cyberattack. This attack somehow affected the page, and it says, “temporarily limiting registrations.” However, the issue has been identified and fixed on January 29. The experts conducted research and estimated that it was a type of denial-of-service (DDoS) attack.

Another concern raised with DeepSeek AI is its privacy and security. It accesses personal information collected and stored by Google. Even its privacy policy says:

“When you use our Services, we may collect your text or audio input, prompt, uploaded files, feedback, chat history, or other content that you provide to our model and Services...the personal information we collect from you may be stored on a server located outside of the country where you live. We store the information we collect in secure servers located in the People's Republic of China.”

So, be aware of sharing any sensitive information, such as name, date of birth, and email address. Furthermore, it also collects technical information, such as IP, operating system, etc. 

Important Person Behind The Success of DeepSeek AI

One of the most important people behind the success of this newly released AI chatbot is Luo Fuli, a 29-year-old AI researcher, also popular as “AI prodigy” in China. Additionally, she worked on natural language processing (NLP) while developing DeepSeek-V2. She studied at Beijing Normal University and later earned a place at Peking University’s Institute of Computational Linguistics. Alibaba and Xiaomi noticed her research skills when she was overcoming major difficulties. She contributed to major research and joined DeepSeek in 2022. 

The Last Line

DeepSeek AI has become popular and surpasses OpenAI’s chatbot and Gemini. Its free-to-use open-source feature is what plays a major role in its popularity. After its release in less than a month, it shocked the US market, which caused Nvidia to lose $500 billion. Furthermore, it is a major competitor to the leading AI chatbots. Luo Fuli, China’s AI prodigy, has made a major contribution as she is responsible for developing natural language processing in DeepSeek V2. 

Frequently Asked Questions (FAQs)

Ques: Who is the owner of DeepSeek?

Ans: The owner of DeepSeek Liang Wenfeng, who is also the owner of the Chinese hedge fund High-Flyer. 

Ques: What is the DeepSeek R1 AI model?

Ans: The DeepSeek R1 AI model is a model that is free to use and allows users to enjoy AI capabilities at minimal cost. 

Ques: How much did DeepSeek cost?

Ans: The development cost of DeepSeek AI is less than $10 million. Moreover, in less than two years, the company gathered 200 employees.