FastAPI with OpenAI Assistant API Integration

Overview

This project demonstrates how to integrate FastAPI with OpenAI's Assistant API, utilizing Server-Sent Events (SSE) for real-time streaming of responses. The application allows creating interactive sessions with an OpenAI Assistant, handling real-time data streaming, and showcasing how asynchronous communication can enhance user interaction.

Now, you can also configure the application to use Azure OpenAI API instead of the standard OpenAI API.

Features

• Assistant Setup: Configures OpenAI Assistant API or Azure OpenAI API based on your choice.
• Thread Management: Create and manage conversation threads.
• Real-time Streaming: Utilize SSE for streaming from the OpenAI Assistant.
• Functionality Extension: Placeholder for future function calling integration.

Getting Started

Prerequisites

• Python 3.10+
• FastAPI
• Uvicorn (for running the application)
• OpenAI Python client

Installation

Clone the repository:

git clone https://github.com/Newpaw/fastapi-assistant-streaming
cd fastapi-assistant-streaming

Install required packages:

pip install -r requirements.txt

Set up environment variables: Create a .env file in the project root directory and add your OpenAI API key and Assistant ID:

For Standard OpenAI:

OPENAI_API_KEY=your_openai_api_key_here
OPENAI_ASSISTANT_ID=your_assistant_id_here
USE_AZURE=True

For Azure OpenAI:

AZURE_OPENAI_API_KEY=your_azure_openai_api_key_here
AZURE_BASE_URL=your_azure_base_url_here
AZURE_OPENAI_ASSISTANT_ID=your_azure_assistant_id_here
USE_AZURE=True

Note: Ensure that USE_AZURE is set to False if you are using OpenAI API. This flag determines which service is utilized.

Obtaining API Keys and Assistant IDs

For Standard OpenAI:

To obtain the necessary credentials, follow these steps:

1. Register and Log In to OpenAI:

   • Go to OpenAI Registration and create an account if you haven't already.
   • After registering, log in to your OpenAI account.

2. Get Your OpenAI API Key:

   • Navigate to the API Keys page.
   • Click on "Create new key" and copy the generated API key. This will be used as your OPENAI_API_KEY.

3. Create an Assistant and Get Assistant ID:

   • Go to the Assistants Page.
   • Follow the instructions to create a new assistant.
   • Once the assistant is created, you'll receive an Assistant ID which you can find in the assistant’s details. This will be used as your OPENAI_ASSISTANT_ID.

For Azure OpenAI:

1. Register and Log In to Azure:
   • Go to Azure Portal and create an account if you haven't already.
2. Create an OpenAI Resource:
   • In the Azure portal, search for "Azure OpenAI" and create a new resource.
   • Once the resource is created, navigate to its key and endpoint page.
3. Get Your Azure OpenAI API Key and Endpoint:
   • From the key and endpoint page, copy your Azure API key and endpoint. Set these as your AZURE_OPENAI_API_KEY and AZURE_BASE_URL in the .env file.
4. Create an Assistant and Get Assistant ID:
   • Similar to standard OpenAI, follow Azure-specific instructions for setting up an assistant and obtaining an Assistant ID.

Make sure to keep your .env file secure and do not share your API keys publicly.

Running the Application

Start the FastAPI Development Server:

uvicorn main:app --reload

Or use dockerfile:

docker build -t fastapi-assistant-streaming .
docker run -p 8000:8000 fastapi-assistant-streaming

or

docker run -p 8000:8000 --env-file .env newpaw/ai-webchat:latest

Testing Endpoints

Check the Health:

curl -X 'GET' --url 'http://localhost:8000/healtcheck'

Get the Assistant:

curl -X 'GET' --url 'http://localhost:8000/api/v1/assistant'

Create a thread:

curl -X POST http://localhost:8000/api/v1/assistant/threads -H "Content-Type: application/json"

Send a message:

curl -N -X POST \
-H "Accept: text/event-stream" -H "Content-Type: application/json" \
-d '{"text": "Hello! Please introduce yourself", "thread_id": "thread_abc123" }' \
http://localhost:8000/api/v1/assistant/chat

License

Distributed under the MIT License. See LICENSE for more information.

Thanks to:

xbreid