# stanfordnlp/stanza

Stanford NLP Python library for tokenization, sentence segmentation, NER, and parsing of many human languages

Repository: https://github.com/stanfordnlp/stanza
Canonical: https://ross.abutalabs.com/products/stanza
Homepage: https://stanfordnlp.github.io/stanza/
Language: Python
License: NOASSERTION
License Family: other
Topics: python, nlp, natural-language-processing, machine-learning, deep-learning, artificial-intelligence, pytorch, universal-dependencies, named-entity-recognition, corenlp
Last push: 2026-08-26T23:51:55+00:00

## Health v2 (maintenance only)
Score: 97/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 93, longevity 100
- inputs: {"age_days": 3263, "days_push": 7, "days_rel": 50, "gap_med": 14, "n_releases_24m": 12}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7867, forks 956 (observed 2026-08-28T04:10:08.336803+00:00)

## What it is
Stanza is the Stanford NLP Group's official Python library for linguistic analysis of human language text. It provides a neural pipeline built on PyTorch for tokenization, sentence segmentation, lemmatization, POS tagging, dependency parsing, and named entity recognition across 70+ languages, plus a Python interface to Java CoreNLP.

## Use cases
- tokenize and segment sentences in many languages
- run named entity recognition on text
- get dependency parses in Universal Dependencies format
- tag parts of speech and morphological features
- extract entities from biomedical literature and clinical notes
- access CoreNLP features like coreference resolution from Python
- train custom NLP models on annotated data

## When to choose
- you need accurate multilingual NLP with pretrained models for 70+ languages
- you want a native Python pipeline without a Java dependency
- you need Universal Dependencies-compliant tokenization and parsing
- you need biomedical/clinical NER models

## When to avoid
- you need fast CPU-only inference for high-throughput production (GPU recommended)
- you want a lightweight rule-based toolkit rather than neural models
- you need tasks like question answering or text generation

## Facets
- artifact type: library
- maturity: stable
- function: nlp, machine-learning, deep-learning, parser
- domain: machine-learning, artificial-intelligence
- platform: python, cross-platform
- tags: pytorch, named-entity-recognition, universal-dependencies, tokenization, dependency-parsing, corenlp, biomedical-nlp, multilingual, natural-language-processing, gpu

## Member repositories
- stanfordnlp/stanza (main) score 97

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:08.336803+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:33:45.383171+00:00, confidence not recorded.
  - readme: https://github.com/stanfordnlp/stanza (fetched 2026-08-28T04:10:08.336803+00:00, sha ff5ca563671c)
  - homepage: https://stanfordnlp.github.io/stanza/ (fetched 2026-08-29T08:30:12.981198+00:00, sha 90ab68be0cd1)
  - registry_pypi: https://pypi.org/pypi/stanza/json (fetched 2026-08-29T08:30:12.983807+00:00, sha 734c7c690ce1)
- Data as of 2026-08-30T08:39:29.467469+00:00.