spaCy

spaCy is an open-source, industrial-strength Natural Language Processing (NLP) library for Python. It is designed for building real-world products and performing large-scale information extraction tasks efficiently.

Features

• Support for 75+ languages
• 84 trained pipelines for 25 languages
• Multi-task learning with pretrained transformers (e.g., BERT)
• Pretrained word vectors
• Linguistically-motivated tokenization
• Components for:
  • Named Entity Recognition (NER)
  • Part-of-Speech (POS) tagging
  • Dependency parsing
  • Sentence segmentation
  • Text classification
  • Lemmatization
  • Morphological analysis
  • Entity linking
  • Span categorization
• Extensible with custom components and attributes
• Support for custom models in PyTorch, TensorFlow, and other frameworks
• Built-in visualizers for syntax and NER
• Easy model packaging, deployment, and workflow management
• Production-ready training system
• Robust and rigorously evaluated accuracy
• State-of-the-art speed
• Large Language Model (LLM) Integration:
  • The spacy-llm package for integrating LLMs into NLP pipelines
  • Modular system for prototyping and prompting
  • Structured outputs from unstructured LLM responses, no training data required
• Reproducible training for custom pipelines
  • Comprehensive configuration system for training runs
  • Easily rerun and track experiments
• End-to-end workflows:
  • Project system for managing data transformation, preprocessing, and training steps
  • Source asset download, command execution, checksum verification, and caching
• Benchmarks:
  • Transformer-based pipelines with state-of-the-art accuracy
  • Multiple pre-trained pipelines with published accuracy metrics on datasets like OntoNotes 5.0 and CoNLL-2003
• Ecosystem:
  • Wide variety of plugins and integrations
  • Community resources, online course, and interactive learning tools

Pricing

spaCy is open-source and free to use.

Tags

python vector-embeddings nlp open-source