Since the "DeepSeek Moment" in January 2025, China's open-source AI ecosystem has evolved significantly. The third part of a three-part series highlights how major Chinese organizations are embracing open source. Companies like Alibaba, Tencent, ByteDance, and Baidu are shifting their strategies. Alibaba has integrated open source into its infrastructure, with its Qwen model becoming a cornerstone for over 113,000 derivatives on Hugging Face. Tencent initially adopted a more closed approach but has since accelerated its open-source releases, particularly with its Tencent Hunyuan models in vision and video. ByteDance is selectively open sourcing components while maintaining a focus on its products, achieving significant scale with its AI application Doubao, which surpassed 100 million daily active users in December 2025. Baidu, known for its reluctance towards open source, is beginning to embrace it with the release of its Ernie 4.5 series and renewed investment in its PaddlePaddle framework. Startups like Moonshot and Z.ai are also contributing to the landscape, with notable models such as Kimi K2 gaining traction. The article emphasizes that the new ecosystem is defined by interconnectedness rather than just an increase in models. Open-source models can be extended and reused, and there's a growing emphasis on infrastructure that supports collaboration. Years of investment in data centers and compute capabilities have laid the groundwork for this transformation. China's compute capacity is projected to grow significantly, with an expected annual increase of 43% for AI-specific resources. This shift reflects a broader trend in which organizations increasingly invest in their own models tailored to specific business needs, rather than relying solely on external providers.