Why China's 2.1 million open source contributors still don't own the top frameworks
As Chinese teams take leading roles in AI infrastructure, Western developers will increasingly run on top of their open source code, making their maintenance habits and community decisions everyone's concern. The article also makes plain that community management and global pain points, not code alone, determine whether an open source project reaches wide adoption.
China's open source participation is huge: 2.1 million active GitHub developers, third globally, and 4.25 million active projects growing 40.58% year-over-year. But the pattern is contributor-heavy, owner-light: Chinese engineers excel at PRs, bug fixes, and docs in other people's repos, yet rarely launch a repo that defines an API for the world. History explains this: top frameworks like React, Kubernetes, Spring, and .NET all grew on platform layers controlled by US companies, so whoever owns the runtime dictates what frameworks can exist.
Silicon Valley runs a virtuous loop where open source builds reputation, which funds full-time maintenance, which strengthens the project. That loop breaks in China: performance reviews ignore star counts, companies veto releases over legal and compliance fears, and many projects die after a burst of conference marketing. The distribution gap is just as wide: clear English READMEs, 24-hour issue responses, roadmaps, and RFC processes are treated as real jobs abroad, but almost no Chinese company staffs them. When technical difference is 20%, distribution difference can be 200%.
The AI infrastructure layer may break the pattern. Six of the ten fastest-growing open source repositories in 2025 are AI infrastructure, and China is already a major player with Qwen, DeepSeek, Dify, RAGFlow, PaddlePaddle, plus heavy contributions to vLLM and llama.cpp. Because this platform is new, it hasn't been locked by two decades of ownership; that's a genuine lane-change opportunity. The author's practical advice: choose problems that hurt developers worldwide, treat English documentation and issue responses as top priority, and build first rather than wait for company approval.
The article reframes 'no open source spirit' as a structural problem: platform ownership, corporate incentives, and distribution capacity matter more than individual effort or wealth.
Because new AI infrastructure lacks entrenched platform owners, Chinese teams have a rare chance to set standards, but they must solve distribution and community management first.
The recommendation to build first and force organizational recognition might work for individuals, but it assumes legal and compliance risks that most corporate employees cannot simply ignore.
Everything domestic is fast-food style—the well-known projects that do come out all have something propping them up.