DeepSeek, a Chinese AI lab, developed an open-source reasoning model called R1 that rivals OpenAI’s proprietary model at a fraction of the cost. Despite going against traditional business practices, DeepSeek chose to open-source their model to build trust in Western markets and navigate export control restrictions on cutting-edge chips. As models become more commoditized, the debate between using open-source versus proprietary software in infrastructure persists. While open-source software is cheaper and customizable, proprietary software offers a smoother experience. However, in infrastructure, open-source solutions are gaining popularity due to their customizability and auditability. OpenAI, despite facing competition from open-source models like R1, remains a leader in the field, pushing for more efficient methods and innovation.
https://www.getlago.com/blog/deepseek-open-source