A Brief History of DeepSeek AI
DeepSeek AI was founded in 2023. The goal was to create an efficient and publicly available AI model for the global community. On January 27, 2025, there were large declines in shares of AI-related technology companies on Wall Street, and their cause was the Chinese startup, which, according to its creators, was created at a fraction of the costs incurred by its US competitors. It is distinguished by its openness - the company decided to make its model available as open source . This means that developers can test and develop this solution without incurring licensing costs.
It's a free AI chatbot that works similarly to ChatGPT. You can use it to perform advertising database the same tasks. The R1 AI model is powerful - it has about 670 billion parameters. It wins in many tests where the capabilities of individual models are compared. This is shown in the graph below:
What is Deepseek AI?
-comparison-openai-chatgpt-google-gemini-meta-llama-rcna189568
According to the creators of the solution, the process of generating answers is similar to the analysis performed by a human. However, it requires less memory than the competition, which is why the costs of the Chinese solution are lower. The entire model was reportedly created for $6 million, which is a very low amount compared to over $100 million in the case of ChatGPT.