LangChain Application Template

Get a /chat API up and running in under 5 minutes with LangChain, FastAPI, and OpenAI.

Getting Started

Prerequisites

• Cookiecutter >= 2.6.0
• uv >= 0.5.11
• OpenAI API Key

Usage

1. From the terminal, run cookiecutter https://github.com/jolfr/langchain-application-template
2. Follow on-screen prompts to configure the project. Paste in the OpenAI API key when prompted.
3. Wait for the project to generate then cd <project_slug> into the project folder.
4. Take a look at the README.md. It has useful commands to help you get up and running.
5. Fire up a development server with uv run bash scripts/dev.sh
6. Navigate to localhost:8000/docs for the Swagger documentation.
7. Test the /chat endpoint by clicking the "Try it out" button and entering a message.

Features

• Service definition with FastAPI
• Preconfigured /chat endpoint
• Basic LangChain chain using OpenAI model provider
• Testing with pytest
• Production builds with Docker
• Environment management with uv