Qwen3-1.7B is a compact, 1.7 billion parameter dense language model from the Qwen3 series, featuring dual-mode operation for both efficient dialogue (non-thinking) and advanced reasoning (thinking). Despite its small size, it supports 32,768-token contexts and delivers strong multilingual, instruction-following, and agentic capabilities, including tool use and structured output.