A compact text model that punches above its weight class for its size. With a 131K token context window, it handles surprisingly long documents for a 1B parameter model. Its small footprint makes it practical for resource-constrained environments, though capability trade-offs compared to larger models are expected.