openAI Discord Bot

What is open AI?

The openAI API is a free to use product that offers developers to utilize ready to go AI. OpenAI offers a multitude of model offering different tasks for use bases. For this project we will utilize the chat GPT-3 and the image generation AI models.

A GPT-3 chatbot is a software application that is able to conduct a conversation with a human user through written or spoken language. The level of “intelligence” among chatbots varies greatly. While some chatbots have a fairly basic understanding of language, others employ sophisticated artificial intelligence (AI) and machine learning (ML) algorithms to achieve an almost human conversational level.

Learn more about openai here

In this tutorial I’m going to show you how easy it is to build and deploy a bot for Discord using the OpenAI platform, Node.js, Repl, and uptimerobot.

Using the chat feature:

Using the image feature:

Create the Node.js implementation

After creating the Discord bot application now we can move on to the more interesting portion.

Before starting this tutorial ensure you have node.js installed on your local machine.

1. Create the Repo

In your chosen CLI run the following:


        mkdir discordBot
        cd discordBot
        npm init

2. Install packages

In your project directory, run the code below to install the necessary modules for our bot development.

npm install discord.js axios dotenv openapi

3. Set up index.js

The index.js will be the driver for most of our code. Within the index we will need to establish connection to the discord bot, the openai, and define logic for listening to user messages.

First set up the required imports by adding this to your index.js


        require('dotenv').config();
        const axios = require('axios');
        const { Client, GatewayIntentBits } = require('discord.js');
        const OpenAI = require('openai');

const keepAlive = require('./server');

Next we can configure the openai connection that our bot will use. You will need an apikey to connect to the service. openai already has a guide on this here


        const {Configuration, OpenAIApi} = OpenAI;
        const configuration = new Configuration({
        apiKey: process.env.OPENAI_API_KEY,
        });
        const openai = new OpenAIApi(configuration);

Notice the process.env.OPENAI_API_KEY? This is set by adding a .env file in your directory. It is good practice to NOT share your apikey with others. .env files help protect your secrets from being exposed at run time. However, while you can use a .env for local testing we will follow a different approach when we deploy to Repl.

Now lets setup how the bot will communicate with the users in Discord



        const client = new Client({ intents: [

        GatewayIntentBits.Guilds,
        GatewayIntentBits.GuildMessages,
        GatewayIntentBits.MessageContent,
        GatewayIntentBits.GuildMembers,

        ], });

Should you have questions on this process please see the Discord documentation.

Now we can add the pieces needed for the chat portion of the AI


        const prefix = "!";

        client.on("messageCreate", async function(message) {

        if (message.author.bot) return;
        if (!message.content.startsWith(prefix)) return;

        const commandBody = message.content.slice(prefix.length);
        var question = commandBody.split(':');
        question.shift();

        console.log(question);
        const args = commandBody.split(' ');
        const command = args.shift().toLowerCase();

        if(command ==="question:"){

        try{

          const response = await openai.createCompletion({
          model: "text-davinci-003",
          prompt: `Pretend you are John who is a Mr know it all. Be mean but answer questions correctly.
          Person: ${message}
          John: `,
          max_tokens: 500,
          temperature: 0,
          });

          console.log(response.data);

          message.reply(response.data.choices[0].text);
          } catch (error) {
          if(error.response){
            console.log(error.response.status);
            console.log(error.response.data);
          } else {
            console.log(error.messageq);
            }
          }
        }

Now lets add the portion needed for generating images from a user request


        if(command ==="image:"){

          try{

            const response = await openai.createImage({
            model: "text-davinci-003",
            prompt: `${message}`,
            n:1,
            size: `1024x1024`,
          });

          console.log(response.data);

          message.reply(response.data.data[0].url);
          } catch(error){
            if(error.response){
              console.log(error.response.status);
              console.log(error.response.data);
            } else {
              console.log(error.messageq);
            }
          }
        }
        });

        keepAlive();

        client.login(process.env.CLIENT_TOKEN);

This is everything that is needed for grabbing a user's message, detecting the question command, passing the question to openai, and returning the response to the user.

4. Set up Server.js

NOTE: If you are going to run locally only and not deploy using Repl do not add the server.js and remove any reference of server.js from the index.js

The server will utilize a function to keep the server alive. Add the below code to your server.js and the code writing portion for a basic chatbot is done!



        const express = require('express');
        const server = express();

        server.all('/', (req, res) =>{

        res.setHeader('Content-Type', 'text/html');
        res.write('<link href="https:fonts.googleapis.com/css?family=?Roboto Condensed"
        rel="stylesheet"> <style> body {font-family: "Roboto Condensed";
        font-size: 22px;}<p>Hosting Active</p>');
        res.end();
        })

        function keepAlive(){
        server.listen(3000, () => {console.log('Server is on')});
        }

        module.exports = keepAlive;

Deploy your bot

With the code in place we can now move to Repl and deploy our code to a live server!

1. Navigate to Repl and create your first account.

2. Create a new project

3. Add all of the files

4. Manage the secrets

Here is where we can enter our openai apikey and our discord bot token. Remember don't share this with anyone!

Make sure you save and you secrets look similar to below

5. Verify the repl output

That is all we needed from repl! At this point you should be able to ping your bot and tak advantage of sending questions to the AI.

Set up Monitoring with UpTimeRobot

To keep Repl from sleeping we will utilize UpTimeRobot to ping our service and keep the bot awake.

You will need to create a free accounthere

1. Hook your site

For this we will ping the repl we deployed every five minutes to determine if the server is awake. Make sure yor https link with the url from repl that ends in 'co'.

Dylan LaMott

Software developer

For the bot source code visit here

Let's Connect!