Large Concept Models (LCM) is a large language model developed by Facebook Research that operates in the sentence representation space, utilizing SONAR embedding to support text in up to 200 languages and speech in 57 languages. LCM is a sequence-to-sequence model designed for autoregressive sentence prediction, exploring various methodologies including mean squared error regression and diffusion-based generative variants. These explorations use a 1.6 billion parameter model trained on approximately 1.3 trillion data points. The main advantages of LCM include its operational capacity for high-level semantic representation and its ability to handle multilingual data. Additionally, LCM's open-source nature allows researchers and developers to access and utilize these models, driving advancements in natural language processing technology.