IllumiNeRF is a 3D relighting technology that recovers a 3D representation from a series of images of an object captured under unknown lighting conditions, allowing it to be rendered from new viewpoints under target illumination. This technique avoids traditional methods based on inverse rendering, which often involve optimization through differentiable Monte Carlo rendering, a process that is both fragile and computationally expensive. IllumiNeRF employs a simpler approach: first, it uses an image diffusion model to relight each input image, and then it uses these re-lit images to reconstruct a Neural Radiance Field (NeRF) model, enabling rendering of new views under target illumination. This approach has achieved surprisingly competitive performance and state-of-the-art results in multiple relighting benchmark tests.