langchain-huggingface

What it is and what it does

langchain-huggingface is a LangChain integration package that bridges Hugging Face models and embeddings into LangChain workflows. It provides classes to work with Hugging Face resources either through local inference or remote API endpoints, letting you embed text and run language models within LangChain applications.

The package has three core runtime dependencies: huggingface-hub (for API access), langchain-core (the LangChain framework), and tokenizers (for text processing). Local inference requires additional optional dependencies (sentence-transformers and transformers), which must be explicitly installed via the [full] extra. The package is actively maintained, supports Python 3.10 through 3.14, and carries no known vulnerabilities.

Use it for:

• Build RAG pipelines that use Hugging Face embeddings to encode documents and queries for semantic search.
• Deploy LangChain agents that call Hugging Face models via API endpoints for inference without local GPU.
• Run local language models through LangChain using Hugging Face transformers for privacy-sensitive applications.
• Combine Hugging Face embeddings with LangChain memory and retrieval for context-aware chatbots.
• Integrate open-source Hugging Face models into existing LangChain workflows without rewriting integration code.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Connects Hugging Face models and embeddings to LangChain applications, providing integration classes for local inference and API-based access to Hugging Face resources.

Yes. This is an active, well-maintained integration for a widely-used ecosystem (Hugging Face). Low install friction, permissive license, no known vulnerabilities, and broad Python version support (3.10–3.14) make it a straightforward choice if you need Hugging Face models or embeddings in LangChain. Be aware that local inference requires the [full] extra with newer dependency versions than langchain-community previously accepted.

Install

langchain-huggingface on PyPI

pip install langchain-huggingface

uv add langchain-huggingface

poetry add langchain-huggingface

Installing langchain-huggingface

Before you install

Low friction install with a pure-Python wheel. Active maintenance with a recent release 120 days ago. Requires Python 3.10 or later. Optional extras (sentence-transformers>=5.2.0, transformers>=5.0.0) needed only for local inference; note that the [full] extra requires newer versions than langchain-community previously accepted.

License in practice

MIT license is permissive and poses no restrictions on commercial or private use.

Quickstart

pip install langchain-huggingface

from langchain_huggingface import HuggingFaceEmbeddings
# or
from langchain_huggingface import HuggingFaceEndpoint

For local inference (HuggingFaceEmbeddings, HuggingFacePipeline), install the [full] extra which adds sentence-transformers>=5.2.0 and transformers>=5.0.0. Requires Python 3.10 or later.

Verify before relying

• Whether the package supports all Hugging Face model types or only a subset
• Performance characteristics when using remote Hugging Face endpoints vs. local inference
• Whether authentication credentials are required for all use cases or only certain endpoints

Package facts

Evidence: langchain_huggingface-1.2.2-py3-none-any.whl

Tags

More Artificial Intelligence packages