CKIPTagger

CKIPTagger is an open-source toolkit for Chinese natural language processing from Academia Sinica’s CKIP Lab in Taiwan. It provides neural models for the three foundational steps of Chinese text analysis — word segmentation, part-of-speech tagging and named-entity recognition — which are notoriously hard for Chinese because written text has no spaces between words. The project has around 1,700 stars and its code was last updated in 2025.

Overview

CKIP has worked on Chinese language technology for decades, and CKIPTagger is its modern neural release: transformer-era models trained on large annotated corpora, packaged as a straightforward Python library. Feed it raw Traditional Chinese sentences and it returns segmented words with POS tags and extracted entities, forming the preprocessing backbone for search, classification and information-extraction pipelines over Chinese text.

Key Features

• Chinese word segmentation, POS tagging and NER in one toolkit
• Neural models from Academia Sinica’s CKIP Lab, leaders in Chinese NLP
• Handles Traditional Chinese text, including Taiwan usage
• Simple Python API: raw sentences in, structured annotations out
• Custom dictionary support to steer segmentation towards your vocabulary
• GPL-3.0 licensed and self-hostable, with models downloadable for offline use

💳 Pricing

Free. CKIPTagger is open source under the GPL-3.0 license; you download the models and run them locally at no cost.

Who Is It For

NLP engineers processing Chinese text, researchers working with Traditional Chinese corpora, and developers building search or extraction features for Chinese-language products.

The toolkit is maintained by the Chinese Knowledge and Information Processing (CKIP) group at Academia Sinica, whose earlier segmentation systems were long a standard reference for Traditional Chinese text processing in Taiwan.