This chapter discusses the launch of Almost, an open and transparent language model by the Allen Institute for AI. It explores the model's features, such as the release of training code, data, and evaluation framework, and raises questions about the increasing trend of openness in open source projects.