A Chinese-language embedding specialist with an unusually wide context window — 1792 dimensions gives it fine-grained representational space for capturing semantic nuance. It handles multilingual retrieval tasks with a lean toward Chinese text, making it practical for search and similarity workflows in that language space. Open weights mean you can run it locally without API dependencies.