# segment-any-text/wtpsplit

Toolkit to segment text into sentences or other semantic units in a robust, efficient and adaptable way.

Repository: https://github.com/segment-any-text/wtpsplit
Canonical: https://ross.abutalabs.com/products/wtpsplit
Language: Python
License: MIT
License Family: permissive
Topics: sentence-boundary-detection, python, deep-learning, machine-learning, pretrained-models, sentence-segmentation, sentence-segmenter, natural-language-processing
Last push: 2026-08-08T22:46:56+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 66, longevity 100
- inputs: {"age_days": 2406, "days_push": 25, "days_rel": 144, "gap_med": 47, "n_releases_24m": 10}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1333, forks 84 (observed 2026-08-28T04:04:24.764940+00:00)

## What it is
wtpsplit is a Python toolkit for segmenting text into sentences or other semantic units using the SaT and WtP deep learning models. It provides robust, punctuation-agnostic sentence segmentation across 85 languages with GPU, TPU, and ONNX inference support.

## Use cases
- split text into sentences without relying on punctuation
- segment multilingual text into sentences across 85 languages
- preprocess text for NLP pipelines like translation or RAG chunking
- adapt sentence segmentation to a specific domain or writing style
- run fast sentence segmentation on GPU or with ONNX

## When to choose
- you need robust sentence segmentation on unpunctuated or noisy text
- you work with many languages and need one universal segmenter
- you need semantic paragraph or section splitting for document chunking
- you want state-of-the-art segmentation performance with GPU/ONNX acceleration

## When to avoid
- you only need simple period-based splitting for clean English text and want zero dependencies
- you cannot run a neural model due to compute or latency constraints
- you need tokenization or full parsing rather than sentence segmentation

## Facets
- artifact type: library
- maturity: active
- function: nlp, machine-learning, parser
- domain: machine-learning, deep-learning
- platform: python, cross-platform
- tags: sentence-segmentation, sentence-boundary-detection, pretrained-models, onnx, multilingual, transformers, natural-language-processing, gpu

## Member repositories
- segment-any-text/wtpsplit (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:24.764940+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 2026-08-30T04:44:55.057520+00:00, confidence not recorded.
  - readme: https://github.com/segment-any-text/wtpsplit (fetched 2026-08-28T04:04:24.764940+00:00, sha 6b3278115677)
  - registry_pypi: https://pypi.org/pypi/wtpsplit/json (fetched 2026-08-29T12:04:05.877532+00:00, sha a3e7215f4948)
- Data as of 2026-08-30T08:39:29.467469+00:00.