Designed for practical, large-scale work, it combines a straightforward API with high processing speed and tools for building production-ready language applications. Its capabilities include support for more than 75 languages and trained pipelines for 25, alongside named entity recognition, part-of-speech tagging, dependency parsing, text classification, sentence segmentation, lemmatization and other linguistic tasks. Developers can use pretrained models and word vectors, integrate transformer models, or extend the library with custom components and models built using frameworks such as PyTorch and TensorFlow.

spaCy also provides visualizers, model packaging and deployment tools, and a configurable training system for reproducible experiments and end-to-end project workflows. Its spacy-llm package helps integrate large language models into structured NLP pipelines. From the makers of spaCy, Prodigy is an annotation tool for tasks such as entity recognition, intent detection and image classification.

The team also offers tailored spaCy pipelines, an interactive online course, documentation, community resources and developer support