GPAvatar is a framework for reconstructing 3D head avatars from images with strong generalization and precise expression control. The method reconstructs 3D head avatars from one or several images in a single forward pass. The key idea is to introduce a dynamic point-based expression field driven by a point cloud to precisely capture expressions. Additionally, a Multi Tri-planes Attention (MTA) fusion module is used in the tri-planes canonical field to leverage information from multiple input images. The proposed method achieves faithful identity reconstruction, precise expression control, and multi-view consistency, demonstrating promising results for free-viewpoint rendering and novel view synthesis. The method is evaluated on the VFHQ and HDTF datasets, showing superior performance in expression control and synthesis quality. The framework is implemented in PyTorch and is available at https://github.com/xg-chu/GPA-avatar. The method is also discussed in terms of ethical implications, including the potential for misuse and the need for watermarks and restrictions on synthetic content. The method is found to be efficient and effective in reconstructing 3D head avatars with high fidelity and precise expression control.GPAvatar is a framework for reconstructing 3D head avatars from images with strong generalization and precise expression control. The method reconstructs 3D head avatars from one or several images in a single forward pass. The key idea is to introduce a dynamic point-based expression field driven by a point cloud to precisely capture expressions. Additionally, a Multi Tri-planes Attention (MTA) fusion module is used in the tri-planes canonical field to leverage information from multiple input images. The proposed method achieves faithful identity reconstruction, precise expression control, and multi-view consistency, demonstrating promising results for free-viewpoint rendering and novel view synthesis. The method is evaluated on the VFHQ and HDTF datasets, showing superior performance in expression control and synthesis quality. The framework is implemented in PyTorch and is available at https://github.com/xg-chu/GPA-avatar. The method is also discussed in terms of ethical implications, including the potential for misuse and the need for watermarks and restrictions on synthetic content. The method is found to be efficient and effective in reconstructing 3D head avatars with high fidelity and precise expression control.