Build Your Own AI Discord Bot with GPT

Want your own AI assistant in Discord? Here's how to build a bot powered by OpenAI's GPT that can answer questions, have conversations, and help your server members.

What We're Building

A Discord bot that:

• Responds to mentions and messages
• Uses GPT-3.5 Turbo for responses
• Maintains conversation context
• Can be customized with personality prompts

Prerequisites

• Discord account and server
• OpenAI API key (pay-as-you-go)
• Node.js 18+ or Python 3.9+
• Basic coding knowledge

Setting Up Discord Application

1. Go to Discord Developer Portal
2. Click "New Application"
3. Name your bot and create
4. Go to "Bot" section
5. Click "Add Bot"
6. Copy the token (keep secret!)
7. Enable required intents:
   • Message Content Intent
   • Server Members Intent

Getting OpenAI API Key

1. Go to OpenAI Platform
2. Create account or sign in
3. Navigate to API Keys
4. Create new secret key
5. Add credits to your account

The Bot Code

Here's a Node.js implementation:

index.js

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

const client = new Client({
  intents: [
    GatewayIntentBits.Guilds,
    GatewayIntentBits.GuildMessages,
    GatewayIntentBits.MessageContent,
  ],
});

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

// Conversation history per channel
const conversations = new Map();

const SYSTEM_PROMPT = `You are a helpful assistant in a Discord server. 
You're friendly, knowledgeable about technology, and can help with coding questions.
Keep responses concise but informative. Use Discord markdown formatting.`;

client.on('messageCreate', async (message) => {
  // Ignore bots and messages without mention
  if (message.author.bot) return;
  if (!message.mentions.has(client.user)) return;

  // Remove the mention from the message
  const userMessage = message.content
    .replace(`<@${client.user.id}>`, '')
    .trim();

  if (!userMessage) {
    return message.reply('How can I help you?');
  }

  // Get or create conversation history
  let history = conversations.get(message.channel.id) || [];
  
  // Add user message to history
  history.push({ role: 'user', content: userMessage });
  
  // Keep last 10 messages for context
  if (history.length > 10) {
    history = history.slice(-10);
  }

  try {
    await message.channel.sendTyping();

    const completion = await openai.chat.completions.create({
      model: 'gpt-3.5-turbo',
      messages: [
        { role: 'system', content: SYSTEM_PROMPT },
        ...history,
      ],
      max_tokens: 500,
    });

    const reply = completion.choices[0].message.content;
    
    // Add assistant response to history
    history.push({ role: 'assistant', content: reply });
    conversations.set(message.channel.id, history);

    // Handle long responses
    if (reply.length > 2000) {
      const chunks = reply.match(/.{1,2000}/g);
      for (const chunk of chunks) {
        await message.reply(chunk);
      }
    } else {
      await message.reply(reply);
    }

  } catch (error) {
    console.error('OpenAI Error:', error);
    await message.reply('Sorry, I encountered an error. Please try again.');
  }
});

client.once('ready', () => {
  console.log(`Logged in as ${client.user.tag}`);
});

client.login(process.env.DISCORD_TOKEN);

Package Configuration

package.json

{
  "name": "discord-gpt-bot",
  "version": "1.0.0",
  "main": "index.js",
  "dependencies": {
    "discord.js": "^14.14.1",
    "openai": "^4.24.1"
  }
}

Environment Variables

.env

DISCORD_TOKEN=your_discord_bot_token
OPENAI_API_KEY=your_openai_api_key

Docker Deployment

docker-compose.yml

services:
  discord-bot:
    build: .
    container_name: discord-gpt-bot
    restart: unless-stopped
    environment:
      - DISCORD_TOKEN=${DISCORD_TOKEN}
      - OPENAI_API_KEY=${OPENAI_API_KEY}

Dockerfile

FROM node:20-alpine

WORKDIR /app
COPY package*.json ./
RUN npm install --production
COPY . .

CMD ["node", "index.js"]

Customization Ideas

Personality Prompts

Make your bot unique:

const SYSTEM_PROMPT = `You are a pirate captain named Captain Byte. 
You speak like a pirate and relate everything to sailing the digital seas.
You help with tech questions but always stay in character.`;

Channel-Specific Behavior

const channelPrompts = {
  'helpdesk': 'You are a helpful support agent...',
  'coding': 'You are a senior developer...',
  'gaming': 'You are a gaming buddy...',
};

Rate Limiting

Prevent API cost overruns:

const cooldowns = new Map();

client.on('messageCreate', async (message) => {
  const cooldown = cooldowns.get(message.author.id);
  if (cooldown && Date.now() - cooldown < 5000) {
    return message.reply('Please wait a few seconds...');
  }
  cooldowns.set(message.author.id, Date.now());
  // ... rest of handler
});

Cost Management

GPT-3.5 Turbo costs approximately:

• Input: $0.0005 per 1K tokens
• Output: $0.0015 per 1K tokens

For a 500-word conversation: ~$0.001

Set usage limits in OpenAI dashboard to avoid surprises.

Learn More

• Discord Bot Documentation
• Discord Comments Integration

---

Built a cool Discord bot? Share it on Discord!