GPT Chatbot

gpt_chatbot is a highly customizable and easy-to-use Flutter package that enables developers to integrate a smart chat interface powered by OpenAI's GPT models into their Flutter applications. The package provides a ready-made chat UI and handles interaction with the GPT API, allowing you to create intelligent conversational agents with minimal effort.

Features

• Customizable Chat Interface: Style the chat bubbles, app bar, and input field to match your app’s theme.
• Real-time Streaming: Stream messages from the AI in real-time, simulating natural typing behavior.
• Auto Scrolling: The chat interface automatically scrolls with new messages for a seamless user experience.
• Easy Integration: Integrate GPT-based chat into your app with minimal code.
• Supports GPT-3.5-turbo and GPT-4o-mini: Designed to work with OpenAI’s GPT-3.5-turbo model by default, with support for other models.
• Flexible Chatbots: Create adaptable chatbots for various applications by setting specific system prompts.

Installation

Add gpt_chatbot to your pubspec.yaml:

dependencies:
  gpt_chatbot: ^0.0.3

Usage

To get started with gpt_chatbot, use the following example code:

import 'package:flutter/material.dart';
import 'package:gpt_chatbot/gpt_chatbot.dart';

class MyChatScreen extends StatelessWidget {
  @override
  Widget build(BuildContext context) {
    final chatGPTService = ChatbotService(
      apiKey: 'your-openai-api-key',
      systemPrompt: 'You are a helpful Dietitian Nutritionist. Provide information about diet, food, workout, and nutritional facts.',
      temperature: 0.5, // Custom temperature setting
      maxTokens: 6000, // Custom max tokens
      model: 'gpt-4o-mini', // Specify a different model if needed
    );

    return ChatbotScreen(
      chatGPTService: chatGPTService,
      appBarTitle: 'Health Chatbot',
      hasBackButton: false,
      appBarBackgroundColor: Colors.pink, // Custom app bar color
      textFieldHint: 'Ask your question...',
      userBubbleDecoration: BoxDecoration(
        color: Colors.pinkAccent, // Custom user bubble color
        borderRadius: BorderRadius.all(Radius.circular(12)),
      ),
      botBubbleDecoration: BoxDecoration(
        color: Color.fromARGB(255, 220, 220, 220), // Custom bot bubble color
        borderRadius: BorderRadius.all(Radius.circular(12)),
      ),
    );
  }
}

Demo

API Documentation

ChatbotService

• ChatbotService: Manages interaction with the OpenAI GPT model.

  Constructor:

  • ChatbotService({required String apiKey, required String systemPrompt, double temperature = 0.5, int maxTokens = 1500, String model = 'gpt-4o-mini'})

  Methods:

  • Stream<String> streamMessages(String message): Streams responses from the GPT model in real-time.

ChatbotScreen

• ChatbotScreen: A widget that provides a chat interface integrated with GPT.

  Constructor:

  • ChatbotScreen({required ChatbotService chatGPTService, String appBarTitle = 'Chat', String textFieldHint = 'Ask something...', BoxDecoration userBubbleDecoration = const BoxDecoration(), BoxDecoration botBubbleDecoration = const BoxDecoration()})

  Parameters:

  • chatGPTService: An instance of ChatbotService.
  • appBarTitle: Title for the AppBar.
  • hasBackButton: Show or hide the back button in the app bar.
  • appBarBackgroundColor: Background color for the app bar.
  • textFieldHint: Hint text for the text field.
  • userBubbleDecoration: Decoration for user messages.
  • botBubbleDecoration: Decoration for bot messages.

Configuration

Configure the ChatbotService with your OpenAI API key and customize the chat behavior by adjusting parameters like systemPrompt, temperature, maxTokens, and model.

If You Find This Package Helpful

Support Me | LinkedIn | X | Have a Project?