Chinese Open Source: A Definitive History

This post originally appeared in Interconnected.

“A big question that’s facing a lot of decision makers is whether Chinese open source becomes a permanent feature of the global technology landscape or a fleeting phenomenon that will be banned and excluded by western capitals.”

Open source used to be a niche topic. China tech also used to be a niche topic. Thanks to AI, the combined topic of “Chinese open source technology” is the topic du jour. It is the story that everyone is trying to understand and make decisions about, from the hacker houses in San Francisco, to the policymaking chambers in Washington, Brussels, New Delhi, and, yes, even Beijing.

I first started writing about Chinese open source over five years ago in this newsletter as a pet topic. Back then, few people in the tech or policy world paid much attention to what Chinese developers were doing on GitHub, the different open source projects and communities that were emerging from tech giants and new upstarts, or why the Chinese government changed its tone about open source from ambivalence and occasional hostility to full embrace as a core part of its national innovation and soft power strategy.

DeepSeek changed everything. Along with Qwen, Kimi, GLM, MiniMax, Stepfun, and even Unitree, Chinese open weight AI models took the world by storm. The traction even earned “open source” a shoutout in the Chinese Premier’s all-important address at the Two Sessions, with an explicit commitment to continue the growth of open source communities and ecosystems as part of the country’s next five-year plan.