Medha (Sanskrit: मेधा, meaning retentive brain power and deep cognitive intelligence) is a repository housing specialized, state-of-the-art reasoning Large Language Models (LLMs). It ingests complex, ambiguous prompts, breaks them down into verifiable logical paths, and formulates structured, multi-step execution plans.