A compact, text-focused embedding specialist built on Qwen3 foundations. At 0.6B parameters, it trades raw capacity for efficiency, producing vector representations of text within a 32K token context window. Its open-weight Apache 2.0 license makes it freely deployable without restrictions.