Titanoboa (; ) is a genus of extinct giant boid snake (being the biological family of all boas and anacondas) that lived during the middle and late Paleocene epoch. Titanoboa was first discovered in the early 2000s by members of the Smithsonian Tropical Research Institute, which–along with students from the University of Florida–recovered 186 fossils of Titanoboa from the Cerrejón coal mines in the La Guajira department of northeastern Colombia. It was named and described in 2009 as Titanoboa cerrejonensis, and lauded as the largest snake ever found at that time. The original type
Titanoboa (; ) is a genus of extinct giant boid snake (being the biological family of all boas and anacondas) that lived during the middle and late Paleocene epoch. Titanoboa was first discovered in the early 2000s by members of the Smithsonian Tropical Research Institute, which–along with students from the University of Florida–recovered 186 fossils of Titanoboa from the Cerrejón coal mines in the La Guajira department of northeastern Colombia. It was named and described in 2009 as Titanoboa cerrejonensis, and lauded as the largest snake ever found at that time. The original type material consisted of thoracic vertebrae and ribs, but later expeditions collected parts of the skull and teeth.
Titanoboa is estimated to grow up to or perhaps even up to long, and weigh around . The discovery of Titanoboa cerrejonensis supplanted the previous record holder, Gigantophis garstini, which is known from the Eocene of Egypt. Titanoboa evolved following the extinction of all nonavian dinosaurs, being one of the largest reptiles that lived after the Cretaceous–Paleogene extinction event. Its vertebrae are very robust and wide, with a pentagonal shape in anterior view, as in other members of Boinae; the genus is considered to be most closely related to other extant boines from the Indo-Pacific region, such as those from Madagascar. Titanoboa is thought to have been a semi-aquatic apex predator, with a diet consisting primarily of fish.
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).