Discord AI Bots

Run multiple AI-powered Discord bots simultaneously, each with their own persona and system prompt. Supports Ollama (local), OpenAI, and Grok as AI providers.

Requirements

• Node.js v18+ or Bun
• A Discord application/bot token for each bot
• An API key for your chosen AI provider (or a running Ollama instance)

Setup

1. Install dependencies

npm install
# or
bun install

2. Create your keys file

Copy the example keys file:

cp keys_ex.json keys.json

Then fill in keys.json with your real tokens:

{
  "discord": {
    "mybotname": "YOUR_DISCORD_BOT_TOKEN"
  },
  "openai": "YOUR_OPENAI_API_KEY",
  "grok": "YOUR_GROK_API_KEY"
}

keys.json is gitignored — never commit it.

3. Create a persona

Add a JSON file to the personas/ folder named after your bot (e.g. personas/mybot.json):

{
  "name": "MyBot",
  "discord_name": "mybotname",
  "system_prompt": "You are MyBot, a helpful assistant.",
  "provider_override": null,
  "model_override": null
}

Field	Description
name	Display name used in conversation history
discord_name	Must match the key used in keys.json under discord
system_prompt	The AI system prompt defining the bot's personality
provider_override	Override the global provider ("ollama", "openai", "grok", or null)
model_override	Override the global model (e.g. "gpt-4o", or null)

4. Register the bot in config.json

Add your bot's name to the bots array in config.json:

{
  "bots": ["mybotname"]
}

The name must match both the persona filename (personas/mybotname.json) and the discord key (keys.json → discord.mybotname).

5. Get your AI provider API key

OpenAI

1. Sign up or log in at platform.openai.com
2. Go to API keys → Create new secret key
3. Copy the key into keys.json under "openai"

Note: OpenAI requires a paid balance to use the API. Add credits at platform.openai.com/settings/billing.

xAI (Grok)

1. Sign up or log in at console.x.ai
2. Go to API Keys → Create API Key
3. Copy the key into keys.json under "grok"

Ollama (free, runs locally)

1. Download and install Ollama from ollama.com/download

2. Pull a model, e.g.:

   ollama pull llama3.2

3. Ollama runs at http://localhost:11434 by default — no API key needed

4. Browse available models at ollama.com/library

---

6. Configure your AI provider

Set provider in config.json to ollama, openai, or grok, and fill in the relevant section:

{
  "provider": "grok",

  "ollama": {
    "base_url": "http://localhost:11434",
    "model": "llama3.2"
  },

  "openai": {
    "base_url": "https://api.openai.com/v1",
    "model": "gpt-4o-mini"
  },

  "grok": {
    "base_url": "https://api.x.ai/v1",
    "model": "grok-4-0709"
  }
}

6. Create a Discord bot application

For each bot:

1. Go to discord.com/developers/applications
2. Click New Application, give it a name
3. Go to Bot → copy the token into keys.json
4. Under Bot, enable Message Content Intent
5. Go to OAuth2 → URL Generator, select bot, then the Send Messages and Read Message History permissions
6. Use the generated URL to invite the bot to your server

7. Run

npm start
# or for auto-reload on file changes:
npm run dev

Configuration reference

All options live in config.json:

Option	Values	Description
provider	ollama, openai, grok	Default AI provider for all bots
bots	array of names	Which bots to start
conversation.mode	mention_only, all_messages	When to respond — only when @mentioned, or to every message
conversation.allowed_channels	array of channel IDs	Restrict bots to specific channels (empty = all channels)
conversation.respond_to_bots	true / false	Whether bots respond to other bots
conversation.max_bot_chain	number	Max consecutive bot messages before going silent (prevents infinite loops)
conversation.max_history	number	How many messages to keep in context per channel
conversation.typing_indicator	true / false	Show "Bot is typing…" while generating a response

Adding a second bot

1. Create another Discord application and copy its token
2. Add the token to keys.json under a new key: "discord": { "alice": "...", "bob": "..." }
3. Create personas/bob.json with its system prompt
4. Add "bob" to the bots array in config.json
5. Restart the process — both bots will run in the same process