The Best 780 Sequence Labeling Tools in 2025

This is a POS tagging model fine-tuned based on the Indonesian RoBERTa model, trained on the indonlu dataset for Indonesian text POS tagging tasks.

BERT fine-tuned named entity recognition model capable of identifying four entity types: Location (LOC), Organization (ORG), Person (PER), and Miscellaneous (MISC)

This model is a sequence labeling model fine-tuned on RoBERTa, designed to identify and remove Protected Health Information (PHI/PII) from medical records.

Flair's built-in fast English 4-class named entity recognition model, based on Flair embeddings and LSTM-CRF architecture, achieving an F1 score of 92.92 on the CoNLL-03 dataset.

French POS tagging model based on Camembert-base, trained using the free-french-treebank dataset

A Spanish named entity recognition model fine-tuned based on the XLM-Roberta-large architecture, with excellent performance on the CoNLL-2002 dataset.

Named entity recognition model fine-tuned on Indonesian NER tasks based on NusaBert-v1.3

Flair framework's built-in large English NER model for 4 entity types, utilizing document-level XLM-R embeddings and FLERT technique, achieving an F1 score of 94.36 on the CoNLL-03 dataset.

A multilingual punctuation prediction model fine-tuned based on xlm-roberta-base, supporting automatic punctuation completion for 12 European languages

Japanese named entity recognition model fine-tuned based on xlm-roberta-base

A fine-tuned version based on GLiNER, optimized for news entity extraction, achieving up to 7.5% higher zero-shot accuracy across 18 benchmark tests

A multilingual model for predicting punctuation in English, Italian, French, and German texts, designed to restore the punctuation structure of transcribed speech.

A multilingual named entity recognition model based on mBERT, supporting 10 high-resource languages, capable of identifying three types of entities: locations, organizations, and person names.

A BERT-large fine-tuned named entity recognition model achieving state-of-the-art performance on the CoNLL-2003 dataset

Flair's standard 4-class French NER model, based on Flair word embeddings and LSTM-CRF architecture, achieves an F1 score of 90.61 on the WikiNER dataset.

A built-in German 4-class large named entity recognition model in the Flair framework, based on XLM-R embeddings and FLERT technique, achieving an F1 score of 92.31 on the CoNLL-03 German dataset.

A named entity recognition model trained on the Romanian NER dataset RONEC based on the xlm-roberta model, achieving an f1-Macro score of 95 on the test set.

A named entity recognition model for 10 high-resource languages, based on a fine-tuned Distil BERT base model, capable of identifying three types of entities: locations, organizations, and persons.

A multilingual named entity recognition model combining neural networks and knowledge bases, supporting 9 languages

An English named entity recognition model fine-tuned on RoBERTa-large, trained on the conll2003 dataset, specifically optimized for entity recognition in email/chat data.

A Named Entity Recognition (NER) model fine-tuned on the wikiner-fr dataset based on camemBERT, excelling in handling named entity recognition tasks in French texts.

A French named entity recognition model fine-tuned based on camemBERT, with added date label functionality

Flair's built-in large English named entity recognition model for 18 classes, trained on the Ontonotes dataset using XLM-R embeddings and FLERT technology.

Flair's built-in English 18-class named entity recognition model, trained on the Ontonotes dataset with an F1 score of 89.27.

A named entity recognition model fine-tuned on the tner/tweetner7 dataset based on roberta-large, specifically designed for entity recognition in Twitter text

State-of-the-art sentence segmentation technology using a 3-layer Transformer architecture, supporting multilingual text segmentation.

Provides Traditional Chinese transformers models and natural language processing tools

Flair's built-in Dutch 4-category named entity recognition large model, based on XLM-R embeddings and FLERT technology, achieves an F1 score of 95.25 on the CoNLL-03 Dutch dataset.

A reading order prediction model that converts text boxes extracted from PDF or detected by OCR into a readable sequence.

This model is used to identify and remove protected health information (PHI/PII) from medical records in compliance with HIPAA privacy standards.

A Russian text named entity recognition (NER) model fine-tuned based on bert-base-multilingual-cased, using BIOLU annotation format, capable of recognizing various entity types such as person names, locations, and organizations.

Flair's built-in standard English 4-category named entity recognition model, based on Flair embeddings and LSTM-CRF architecture, achieving an F1 score of 93.06 on the CoNLL-03 dataset.

Piiranha-v1 is a fine-tuned model based on microsoft/mdeberta-v3-base, specifically designed to detect 17 types of personally identifiable information (PII) in six languages.

A Turkish named entity recognition model fine-tuned based on the dbmdz/bert-base-turkish-cased model, capable of recognizing entities such as person names, organization names, and location names.

An English named entity recognition model trained on distilbert-base-uncased, specifically designed for identifying biomedical entities (107 entity categories), suitable for text corpora such as case reports.

A BPMN process information extraction model fine-tuned based on bert-base-cased, used to extract key elements such as executors and tasks from textual process descriptions

A tool for restoring punctuation marks in Chinese Wikipedia texts, supporting the restoration of 6 common punctuation marks.

Named entity recognition model fine-tuned on the English CoNLL2003 dataset based on XLM-RoBERTa-large

A model for Chinese part-of-speech tagging, supporting 17 common POS tags.

A fine-tuned version based on the Spanish BERT cased model (BETO) on the NER-C dataset, specifically designed for Named Entity Recognition (NER) tasks.

A Transformer-based Persian language understanding model, reconstructed vocabulary and fine-tuned on new corpora, expanding multi-domain application capabilities

GLiNER is a Named Entity Recognition (NER) model capable of identifying various types of Personally Identifiable Information (PII).