# hezarai/hezar

The all-in-one AI library for Persian, supporting a wide variety of tasks and modalities!

Repository: https://github.com/hezarai/hezar
Canonical: https://ross.abutalabs.com/products/hezar
Homepage: https://hezarai.github.io/hezar/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: persian, persian-nlp, persian-ai, persian-speech-recognition, hezar, persian-ocr, hezarai, persian-image-captioning, persian-dataset
Last push: 2026-03-09T12:52:24+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 71, release rhythm 74, longevity 100
- inputs: {"age_days": 1567, "days_push": 177, "days_rel": 177, "gap_med": 0.5, "n_releases_24m": 9}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1013, forks 65 (observed 2026-08-28T04:03:13.568304+00:00)

## What it is
Hezar is an all-in-one Python AI library for the Persian language, covering NLP, speech recognition, OCR, and image captioning through a task-based model interface. It integrates tightly with the Hugging Face Hub and bundles datasets, embeddings, tokenizers, and training tools.

## Use cases
- persian text classification and sentiment analysis
- persian speech recognition from audio
- persian ocr on scanned documents
- persian named entity recognition and pos tagging
- persian image captioning
- train or fine-tune models on persian datasets
- use persian word embeddings like fasttext and word2vec

## When to choose
- you need ready-to-use pretrained models for Persian across multiple modalities
- you want a simple task-based API instead of wiring transformers pipelines yourself
- you need Persian datasets, metrics, and training recipes in one package

## When to avoid
- your project targets languages other than Persian
- you need cutting-edge custom model architectures not included in Hezar
- you want a lightweight dependency - Hezar pulls in deep learning frameworks

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, nlp, speech-recognition, ocr, image-processing, llm-training, transformers
- domain: machine-learning, speech-processing, image-processing
- platform: python, cross-platform
- tags: persian, persian-nlp, hugging-face, farsi, multimodal, pretrained-models, natural-language-processing

## Member repositories
- hezarai/hezar (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:13.568304+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 2026-08-30T07:11:39.738401+00:00, confidence not recorded.
  - readme: https://github.com/hezarai/hezar (fetched 2026-08-28T04:03:13.568304+00:00, sha 85663a1560ec)
  - homepage: https://hezarai.github.io/hezar/ (fetched 2026-08-29T13:11:12.334406+00:00, sha 438406d195c2)
  - registry_pypi: https://pypi.org/pypi/hezar/json (fetched 2026-08-29T13:11:12.337284+00:00, sha 8022e00553a3)
- Data as of 2026-08-30T08:39:29.467469+00:00.