How a top Chinese AI model overcame US sanctions January 24, 2025 By admin With a new reasoning model that matches the performance of ChatGPT o1, DeepSeek managed to turn restrictions into innovation.