🦜🔗 Databricks LangChain Integration

The databricks-langchain package provides seamless integration of Databricks AI features into LangChain applications. This repository is now the central hub for all Databricks-related LangChain components, consolidating previous packages such as langchain-databricks and langchain-community.

Installation

From PyPI

pip install databricks-langchain

From Source

pip install git+https://git@github.com/databricks/databricks-ai-bridge.git#subdirectory=integrations/langchain

Key Features

• LLMs Integration: Use Databricks-hosted large language models (LLMs) like Llama and Mixtral through ChatDatabricks.
• Vector Search: Store and query vector representations using DatabricksVectorSearch.
• Embeddings: Generate embeddings with DatabricksEmbeddings.
• Genie: Use Genie in Langchain.
• Deep Agents Sandbox: Run Deep Agents shell and filesystem tools inside a Databricks sandbox with DatabricksSandbox (install databricks-langchain[deepagents]).
• Sandbox Tool: Run shell commands in a Databricks sandbox from any LangChain or LangGraph agent with DatabricksSandboxTool (install databricks-langchain[sandbox]).
• Sandbox Session: Provision, wait for, and auto-delete a sandbox for that tool with the DatabricksSandboxSession context manager, keeping sandbox lifecycle in your application rather than the model.

Getting Started

Use LLMs on Databricks

from databricks_langchain import ChatDatabricks

llm = ChatDatabricks(endpoint="databricks-meta-llama-3-1-70b-instruct")

Use a Genie Space as an Agent (Preview)

Note: Requires Genie API Private Preview. Contact your Databricks account team for enablement.

from databricks_langchain.genie import GenieAgent

genie_agent = GenieAgent(
    "space-id", "Genie",
    description="This Genie space has access to sales data in Europe"
)

Contribution Guide

We welcome contributions! Please see our contribution guidelines for details.

License

This project is licensed under the MIT License.

Thank you for using Databricks LangChain!