DeepSeek
Chinese AI company whose open-weights reasoning model R1 (2025) rivaled Western frontier systems at a reported fraction of the training cost and shook markets.
DeepSeek is a Chinese AI company that drew global attention in January 2025 with R1, an open-weights reasoning model that matched systems such as OpenAI o1 on many public tests while claiming much lower training cost.
R1 used reinforcement learning, long chains of thought, and a mixture-of-experts design. The release strengthened the open-weights camp and briefly hit chip-maker valuations, sharpening debates about how much compute frontier LLMs really need. See the article and timeline.