A comprehensive survey on 3D content generation explores recent advancements in artificial intelligence generated content (AIGC), focusing on 3D generation techniques. The paper introduces a new taxonomy categorizing methods into 3D native, 2D prior-based, and hybrid approaches. It reviews approximately 60 papers, discussing technical challenges and future directions. 3D native methods generate 3D content directly from 3D data, while 2D prior-based methods use 2D diffusion models to generate 3D content. Hybrid methods combine both approaches. The survey highlights the importance of 3D representation, including explicit and implicit methods, and discusses the role of 2D diffusion models in 3D generation. Key challenges include generating high-quality 3D content, ensuring consistency across views, and improving generation speed. The paper also addresses data limitations, model architecture improvements, and the need for robust evaluation metrics. Future directions include enhancing 3D generation quality, expanding data sources, and developing more efficient and controllable generation methods. The survey provides a comprehensive overview of current techniques and identifies key areas for further research in 3D content generation.A comprehensive survey on 3D content generation explores recent advancements in artificial intelligence generated content (AIGC), focusing on 3D generation techniques. The paper introduces a new taxonomy categorizing methods into 3D native, 2D prior-based, and hybrid approaches. It reviews approximately 60 papers, discussing technical challenges and future directions. 3D native methods generate 3D content directly from 3D data, while 2D prior-based methods use 2D diffusion models to generate 3D content. Hybrid methods combine both approaches. The survey highlights the importance of 3D representation, including explicit and implicit methods, and discusses the role of 2D diffusion models in 3D generation. Key challenges include generating high-quality 3D content, ensuring consistency across views, and improving generation speed. The paper also addresses data limitations, model architecture improvements, and the need for robust evaluation metrics. Future directions include enhancing 3D generation quality, expanding data sources, and developing more efficient and controllable generation methods. The survey provides a comprehensive overview of current techniques and identifies key areas for further research in 3D content generation.