Two years ago, we successfully introduced the '3DTeethSeg' challenge dealing with teeth segmentation and labeling tasks from intraoral 3D scans. Continuing from our previous challenge and striving for in-depth perception of intraoral scans, we intend to address within this version of the challenge a more complex task, teeth landmark detection. This task holds significant importance in modern clinical orthodontics. These crucial landmarks, including features such as cusps and mesial-distal locations, play a fundamental role in advancing orthodontic treatment planning and assessment in clinical dentistry. However, several significant challenges could be present given the intricate geometry of individual teeth and substantial variations between individuals. To address these complexities, the development of advanced techniques, particularly through the application of deep learning, is required for the precise detection of 3D tooth landmarks. This challenge introduces the first publicly available dataset for 3D teeth landmarks detection, encouraging community involvement in a topic with important clinical implications. It plays a key role in advancing automation and leveraging AI for optimizing orthodontic treatments.