Reports and analyses from leading AI research institutions

DeepSeek-R1: Chinese AI Startup's Open-Source Model Challenges OpenAI Using Reinforcement Learning

Jan 26, 2025, 09:38 AM

Chinese AI startup DeepSeek has made significant strides in the artificial intelligence sector with the release of its open-source large language model, DeepSeek-R1. This model has garnered attention for its ability to match or exceed the performance of OpenAI's o1 model in various benchmarks, including mathematics and coding, at a fraction of the cost. Developed using reinforcement learning, a method that allows the model to learn through trial and error, DeepSeek-R1's development cost was reported to be as low as $5.6 million, significantly less than the costs associated with training comparable models by leading U.S. companies. DeepSeek's success has sparked discussions in Silicon Valley about the potential of open-source models to challenge the dominance of proprietary systems. The company's approach has been praised for its efficiency and transparency, with some industry leaders suggesting that it could influence future AI development strategies globally.