A compact, lightweight model that prioritizes accessibility over raw power. At 150M parameters, it operates within tight resource constraints, making it suited for environments where memory and compute are limited. Its small context window of 2048 tokens means it works best with focused, concise inputs rather than lengthy documents.