Repolode
Submit
stanfordnlp

stanfordnlp/stanza

Stanford NLP Group's Python library for tokenization, sentence segmentation, NER, and dependency parsing across 60+ human languages.

View on GitHubWebsite
Stars
7,906
Forks
966
Open issues
103
Last commit
2 days ago

What is it?

Stanza is a Python natural language processing library from the Stanford NLP Group. It runs neural pipelines for tokenization, sentence segmentation, multi-word token expansion, part-of-speech and morphological tagging, lemmatization, dependency parsing and named entity recognition on text in more than 60 languages.

Beyond the neural pipeline, Stanza includes a wrapper that lets Python code call the Java Stanford CoreNLP software. Users point the wrapper at a local CoreNLP installation through the CORENLP_HOME environment variable and extract annotations from it. Pretrained models cover all Universal Dependencies treebanks in version 2.8, along with NER models for several widely spoken languages and biomedical and clinical English packages.

The library is built on PyTorch and installs via pip or conda. Developers can train their own tokenizer, tagger, lemmatizer, parser or NER models from CoNLL-U or BIOES data by running the training code from the source repository.

Key features

  • 01Neural pipelines for tokenization, POS tagging, lemmatization, and dependency parsing
  • 02Named entity recognition models for several widely spoken languages
  • 03Pretrained models for all Universal Dependencies 2.8 treebanks
  • 04Python wrapper for calling the Java Stanford CoreNLP software
  • 05Biomedical and clinical English model packages
  • 06Train custom models from CoNLL-U or BIOES data

Use cases

  • Parsing and annotating text in many languages within a Python NLP pipeline
  • Extracting named entities from biomedical literature or clinical notes
  • Running Stanford CoreNLP annotators from Python code
  • Training a custom parser or tagger for a low-resource language treebank

Quick start

  1. 1

    Install Stanza with pip

    pip install stanza
  2. 2

    Start Python and import Stanza, then download English models (optional)

    python -c "import stanza; stanza.download('en')"
  3. 3

    Create an English neural pipeline and parse a sentence

Things to know

  • Training your own models requires cloning the repository, since the Pipeline interface does not support training.
  • Using the Java CoreNLP wrapper requires separately downloading Stanford CoreNLP and its models.
  • Anaconda installation does not work for Python 3.10; use pip instead.
  • The license is listed as NOASSERTION in metadata, though the README states Apache License 2.0.

In collections

Natural language processing tools & libraries

Topics

#nlp#python#pytorch#named-entity-recognition#dependency-parsing#tokenization#multilingual#universal-dependencies#artificial-intelligence#corenlp

FAQ

What is stanza?
Stanford NLP Group's Python library for tokenization, sentence segmentation, NER, and dependency parsing across 60+ human languages. Stanza is a Python natural language processing library from the Stanford NLP Group. It runs neural pipelines for tokenization, sentence segmentation, multi-word token expansion, part-of-speech and morphological tagging, lemmatization, dependency parsing and named entity recognition on text in more than 60 languages.
Is stanza free to use? What license does it have?
stanza is open source under the Other license. Check the license terms before using it in commercial products.
Is stanza still maintained?
Actively maintained. The last commit to stanza was on Nov 5, 2026.
What can I use stanza for?
Parsing and annotating text in many languages within a Python NLP pipeline Extracting named entities from biomedical literature or clinical notes Running Stanford CoreNLP annotators from Python code Training a custom parser or tagger for a low-resource language treebank
How do I install stanza?
Install Stanza with pip: pip install stanza Start Python and import Stanza, then download English models (optional): python -c "import stanza; stanza.download('en')" Create an English neural pipeline and parse a sentence
Where can I get stanza?
The source code and installation instructions are on GitHub: https://github.com/stanfordnlp/stanza

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects