{"categories":[{"label":"Libraries","url":"https://skillfed.io/packages/category/software-development-libraries/4"},{"label":"Artificial Intelligence","url":"https://skillfed.io/packages/category/scientific-engineering-artificial-intelligence/4"}],"enrichment":{"capability":"Model2Vec converts sentence transformers into small, fast static embedding models that generate vector representations of text for tasks like retrieval, classification, and clustering.","skillfed_tags":["embeddings","model-distillation","nlp"],"use_cases":["Build a semantic search or retrieval system where inference speed and model size are critical constraints.","Distill a custom static embedding model from a sentence transformer in under a minute without training data.","Fine-tune a classification model on top of a pre-trained Model2Vec embedding for text categorization tasks.","Deploy embeddings in resource-constrained environments where model size and CPU inference speed matter.","Generate multilingual embeddings for text in any of 101 languages using the potion-multilingual model."],"what_it_does":"Model2Vec is a distillation technique that transforms any sentence transformer into a compact static embedding model. It reduces model size by up to 50 times and achieves up to 500 times faster inference on CPU, with minimal performance loss. The package provides pre-trained models from HuggingFace (including multilingual variants) ready for immediate use, plus tools to distill your own models from existing sentence transformers in about 30 seconds without requiring a dataset.\n\nThe core workflow is straightforward: load a pre-trained Model2Vec model or distill one from a sentence transformer, then call encode() to generate sentence embeddings or encode_as_sequence() for token-level embeddings. These embeddings work for text classification, semantic search, clustering, and retrieval-augmented generation. The package integrates with HuggingFace hub for easy model sharing and is already integrated into Sentence Transformers and LangChain.","worth_installing":"Yes. Model2Vec is actively maintained, has no known vulnerabilities, uses permissive MIT licensing, and offers a clear value proposition: fast, small static embeddings with strong performance. Install if you need embedding inference speed and model compactness; the low dependency footprint and HuggingFace integration make it straightforward to adopt. Requires Python >=3.10."},"id":"model2vec","links":{"html":"https://skillfed.io/packages/model2vec","md":"https://skillfed.io/packages/model2vec.md","pypi":"https://pypi.org/project/model2vec/"},"maintenance":{"status":"active"},"meta":{"latest_release":"2026-08-12","license_spdx":null,"license_treatment":"permissive","name":"model2vec","python_support":"supports_current","summary":"Fast State-of-the-Art Static Embeddings"},"popularity":{"monthly_downloads":941618,"position":4676,"tier":"top_5000"},"security":{"n_vulnerabilities":0},"version":"0.9.0"}