China's government aims to become an innovation nation and promotes the development of so‐called indigenous innovation. Under this paradigm of state‐encouraged innovation, however, it is unclear how domestic firms organize their innovation processes. We distinguish between two strategies in that respect: closed versus open innovation. Our findings suggest that firms with closed innovation processes collaborate in close geographic distance, rely on DUI‐modes of learning, and collaborations are based on guanxi. In contrast, firms with open innovation processes collaborate over large distances and rely on STI‐modes of learning that are not necessarily guanxi‐based. The findings help to understand the heterogeneous nature of indigenous innovation in China.