# roshan-research/hazm

Persian NLP Toolkit

Repository: https://github.com/roshan-research/hazm
Canonical: https://ross.abutalabs.com/products/hazm
Homepage: https://www.roshan-ai.ir/hazm/
Language: Python
License: MIT
License Family: permissive
Topics: natural-language-processing, python, persian, nlp, persian-nlp, dependency-parser, embeddings, text-processing, tokenizer, lemmatization, farsi, normalization, pos-tagging
Last push: 2026-04-01T18:55:05+00:00

## Health v2 (maintenance only)
Score: 76/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 75, release rhythm 65, longevity 100
- inputs: {"age_days": 4691, "days_push": 154, "days_rel": 154, "gap_med": 51.0, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1417, forks 208 (observed 2026-08-28T04:04:40.060375+00:00)

## What it is
Hazm is a Python library for natural language processing on Persian (Farsi) text, offering normalization, tokenization, stemming, lemmatization, POS tagging, dependency parsing, and embeddings. It integrates with Hugging Face Hub to automatically download and cache pretrained models.

## Use cases
- normalize persian text
- tokenize farsi sentences and words
- lemmatize persian verbs and nouns
- pos tagging for persian text
- dependency parsing of persian sentences
- create persian word embeddings
- read persian corpora in python

## When to choose
- you need Persian-specific NLP preprocessing in Python
- you want pretrained Persian models with automatic Hugging Face download
- you need to read popular Persian corpora with ready-made scripts

## When to avoid
- you need multilingual NLP beyond Persian
- you work with Python versions older than 3.12
- you need a full end-to-end NLP pipeline with training tooling

## Facets
- artifact type: library
- maturity: active
- function: nlp, parser, machine-learning, search-engine
- domain: machine-learning, developer-tools
- platform: python, cross-platform
- tags: persian, farsi, tokenization, lemmatization, pos-tagging, dependency-parsing, embeddings, text-normalization, hugging-face, natural-language-processing

## Member repositories
- roshan-research/hazm (main) score 76

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:40.060375+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 2026-08-30T04:38:02.239489+00:00, confidence not recorded.
  - readme: https://github.com/roshan-research/hazm (fetched 2026-08-28T04:04:40.060375+00:00, sha 9ec4a803f825)
  - homepage: https://www.roshan-ai.ir/hazm/ (fetched 2026-08-29T11:50:42.951185+00:00, sha 0950a86535c6)
  - registry_pypi: https://pypi.org/pypi/hazm/json (fetched 2026-08-29T11:50:42.960257+00:00, sha 6515e8f075c6)
- Data as of 2026-08-30T08:39:29.467469+00:00.