Langchain

Integrate Mixedbread's powerful embedding and reranking capabilities into your LangChain projects. This guide covers installation, quick start examples for both Python and TypeScript, and links to detailed documentation for seamless integration with your natural language processing workflows.

Integrate Mixedbread's reranking and embeddings capabilities into your LangChain projects.

Quick Start

Install the package:

Installation

pip install langchain-mixedbread-ai

Set up your API key:

export MXBAI_API_KEY=your_api_key_here

Start using Mixedbread in your LangChain projects!

Embeddings

Generate text embeddings for queries and documents.

from langchain_mixedbread_ai import MixedbreadAIEmbeddings
 
embeddings = MixedbreadAIEmbeddings()
 
# Single query
query_embedding = embeddings.embed_query("Is baking fun?")
print(query_embedding)
 
# Multiple documents
documents = ["Baking bread is fun", "I love baking"]
doc_embeddings = embeddings.embed_documents(documents)
print(doc_embeddings)

Learn more about Mixedbread Embeddings

Reranking

Reorder documents based on relevance to a query.

from langchain_mixedbread_ai import MixedbreadAIReranker
 
reranker = MixedbreadAIReranker()
query = "What ingredients are needed for bread?"
documents = [
    "Flour is essential for baking bread",
    "Yeast helps bread rise",
    "Bread tastes great with butter",
]
 
result = reranker.compress_documents(documents, query)
 
print(result)

Learn more about Mixedbread Reranking

Documentation

For detailed information on using Mixedbread with LangChain, check out the LangChain docs directly.

Need Help?

Join our Discord community
Contact support