Skip to main content
Open on GitHub

huggingface

title: Hugging Face Integration description: Learn how to use Hugging Face models, embeddings, and tools with LangChain.​

🚀 Hugging Face Integration in LangChain

All functionality related to the Hugging Face Platform is available through the langchain-huggingface and langchain-community packages.

📦 Installation​

Install the Hugging Face integration packages:

pip install langchain-huggingface

To use tools, embeddings, and TTS (Text-to-Speech), install:

pip install transformers datasets huggingface_hub pillow

🤖 Chat Models​

ChatHuggingFace​

Use chat models via Hugging Face’s hosted models or APIs.

from langchain_huggingface import ChatHuggingFace
API Reference:ChatHuggingFace

📄 Usage Example

💬 LLMs (Text Generation)​

HuggingFaceEndpoint​

Use models directly from Hugging Face Hub (via endpoint):

from langchain_huggingface import HuggingFaceEndpoint

llm = HuggingFaceEndpoint(
    repo_id="tiiuae/falcon-7b-instruct",
    task="text-generation"
)
API Reference:HuggingFaceEndpoint

📄 Usage Example

HuggingFacePipeline​

Run models locally via Hugging Face’s transformers.Pipeline:

from langchain_huggingface import HuggingFacePipeline
API Reference:HuggingFacePipeline

📄 Usage Example

🧠 Embedding Models​

HuggingFaceEmbeddings​

from langchain_huggingface import HuggingFaceEmbeddings
API Reference:HuggingFaceEmbeddings

📂 Document Loaders​

HuggingFaceDatasetLoader​

Load public datasets from Hugging Face Hub.

from langchain_community.document_loaders.hugging_face_dataset import HuggingFaceDatasetLoader

📄 Usage Example

HuggingFaceModelLoader​

Load model metadata and README from Hugging Face Hub.

from langchain_community.document_loaders import HuggingFaceModelLoader

ImageCaptionLoader​

Use Hugging Face vision models to generate image captions.

from langchain_community.document_loaders import ImageCaptionLoader
API Reference:ImageCaptionLoader

📦 Requires:

pip install transformers pillow

🛠️ Tools​

Hugging Face Hub Tools​

Interact with Hugging Face tools directly using:

from langchain_community.agent_toolkits.load_tools import load_huggingface_tool
API Reference:load_huggingface_tool

📦 Requires:

pip install transformers huggingface_hub

📄 Usage Example

HuggingFaceTextToSpeechModelInference​

Wrapper around text-to-speech model using Hugging Face API:

from langchain_community.tools.audio import HuggingFaceTextToSpeechModelInference

💡 For more examples and updates, refer to the official LangChain Hugging Face docs.

Was this page helpful?