INSTA creates instant volumetric head avatars by combining neural networks with volumetric rendering. Built on instant-ngp and the FLAME 3D face model, it reconstructs animatable digital avatars from input images or video sequences. The method leverages CUDA acceleration and neural radiance fields to produce high-quality facial reconstructions that can be used for avatar animation and telepresence applications.