A focused code-generation specialist that handles programming tasks with a 8k context window. It produces code completions, explanations, and instruction-following responses across common programming languages. As a 34B parameter model, it balances capability with the resource requirements that come with running larger open-weight models locally.