Originally made by MishalHossin and all credit for this excellent code goes to them.
JOIN THE DISCORD! To see an example bot in action, for support with installation, to share what you've created, and to hang out with cool people.
Designed With Linux In Mind!
- Free LLM Model: Use this powerful language model without spending anything (for text responses; voice responses require an ElevenLabs subscription).
- Mention Recognition: The bot will always respond when mentioned or called by name, as well as when replied to using Discord's reply feature.
- Message Handling: The bot knows when you're replying to someone else, avoiding confusion.
- Channel-Specific Responses: Use the /toggleactive command to disable the bot in specific channels.
- Web Access: Internet-enabled LLM to allow for current awareness.
- Slash commands: Use /help for a list of all commands that are available on the bot.
- Tiered slash commands: Only people with the ability to manage the guild can use the /enable_voice and /disable_voice functions
- NEW! Voice-Enabled: The bot will check to see if the author is in a voice channel. If they are, the bot will attempt to join and speak the text using ElevenLabs API. (An ElevenLabs subscription is required for this.)
- Toggleable voice functions: /enable_voice and /disable_voice can be used by people with manage server permissions to enable and disable the voice functions, if you do not wish to use them/pay for ElevenLabs.
- NEW! Interjections: If the bot hasn't been triggered in awhile, it will interject randomly in any text channel that it has permission to type in with a context-aware AI generation much the same as any other chat that the bot might normally reply with.
- These interjection chances are adjustable within
config.ymland after each interjection there is a customizable cooldown period (default: 10min) before it will begin considering an interjection again. The interjections are slash-command enabled, meaning you can disable them with /diable_interjections or enable them with /enable_interjections. /interjection_status will show the current status of interjections. As with other slash commands, these may only be used by users with manage server permissions.
- These interjection chances are adjustable within
- Consistently updated: This project is not dead! I contribute to this project regularly with updates, and is designed with Linux and security in mind.
Language Selection 🌐⚙️ (Click to Expand)
To select a Language, set the value of "LANGUAGE" of config.yml with the valid Language Codes listed below:
en- English 🇺🇸de- Deutsch 🇩🇪tr- Türkçe 🇹🇷ar- Arabic 🇦🇪fr- Français 🇫🇷es- Español 🇪🇸vn- Vietnamese 🇻🇳cn- Chinese 🇨🇳ru- Russian 🇷🇺ua- Ukrainian 🇺🇦pt- Português 🇧🇷pl- Polish 🇵🇱
To create a custom personality, follow these steps:
- Create a
.txtfile likeyourfile.txtinside theinstructionsfolder. - Add the way you want the bot to act in
yourfile.txt - Open the
config.ymlfile and locate line 17. - Set the value of INSTRUCTIONS at line 17 as
"yourfile"to specify the custom persona.
yourfile for persona name and set it in config.yml, name it whatever you want as long as you update config.yml to match.
Follow this guide in its entirety. Or don't, I'm not your real dad. If you run into problems, use the internet or ChatGPT - they are your friends and ChatGPT is actually surprisingly good at interpreting code errors if you don't know how.
- This installation guide and all writings on this page were designed with Linux in mind. The tutorial assumes you are using the CLI to operate. Linux was used to create/modify this bot and so all commands listed here will be in Linux format. If you are compiling/editing this bot on Windows, you may not be able to copy/paste all the below commands.
- This bot runs on Python <3.13. If you try to run this bot on Python 3.13, it will fail and you will be confused. This bot MUST use Python 3.12 or older. Use venv or pipx to install older python versions and make sure that to run the bot you use python3.12 main.py and not python main.py.
git clone https://github.com/ArchAngel2190/Discord.Bot/
cd Discord.Bot
Ensure that if you have Python 3.13 or newer installed that you are operating in a virtual environment (venv or pipx) with Python 3.12 or older. If you have multiple versions of Python installed, use the below code except with python3.12 as the leading.
python3 -m pip install -r requirements.txt
FFMpeg must be installed for voice functionality. This is in requirements.txt but may not install properly through that, so use the below commmand to ensure it is properly installed. (Only required on the host machine, not on that of everyone who uses it.)
sudo apt-get install ffmpeg
Step 4. Get your Discord bot token and enable intents from HERE
The intents the bot needs are Presence, Server Members, and Message Content. These should be enabled by default.
Copy your bot's secret token (only shown once, so make sure you copy it then or you will have to regenerate it)
Step 5. Create a Groq account
Step 6. Get Groq api key
Step 8. Remove the prefix from remove_this_prefix.env so the file is titled .env. If you are on Linux, this will cause the file to be hidden in a file manager, so know how to find it first if you're not using the CLI like a Chad.
When properly filled out, .env will look like this:
DISCORD_TOKEN=DISCORD_TOKEN_HERE
API_KEY=GROQ_TOKEN_HERE
ELEVENLABS_API_KEY=ELEVENLABS_TOKEN_HERE
ELEVENLABS_VOICE_ID=VOICE_ID_HERE
Go to Discord.Bot/Instructions and create a text file that describes how you want the bot to act. Use plain English, this isn't rocket science - just tell it who it is/what to do. Use the files inside as an example but you can do basically whatever you want here. This is great for making characters.
python main.py
Ensure you use whatever Python version you have installed. If you are not using a virtual environment for older Python, you might have to use the command pythonx.xx in place of python where x.xx is your Python version. (This assumes you have that version of Python installed).
You may need to run as admin if you are on Windows (but you shouldn't be - why do that to yourself?)
You can Invite your bot using the link in console, or by creating an invite link in Discord Developer's console. Inviting from the generated link allows you to use the bot without its own role.
- Invite your bot and DM (Direct Message) it |
⚠️ Make sure you have DM enabled - Mention (ping) or reply to the bot in any text channel that it has permission to type in.
- If the bot has gone awhile without being pinged (default: 10mins) and a conversation is continuing in the chat without the bot, the bot will interject into the conversation with a context-aware comment.
- Have a working bot token
- Follow up to step 9
For Debian-based distributions (such as Ubuntu):
apt update -y; sudo apt upgrade -y; sudo apt autoremove -y; sudo apt install docker-compose -y
Other Linux distro (Click to expand)
For Red Hat-based distributions (such as CentOS and Fedora):
sudo yum update -y && sudo yum install -y docker-compose
For Arch-based distributions (such as Arch Linux):
sudo pacman -Syu --noconfirm && sudo pacman -S --noconfirm docker-compose
For SUSE-based distributions (such as openSUSE):
sudo zypper update -y && sudo zypper install -y docker-compose
sudo docker-compose up --build
This entire bot was built by MishalHossin and I take no credit for its creation.
Certain parts of their code were modified to create this bot in its present form.
- Updated bot to stop using Asynchronus DuckDuckGoSearch as this is not compat with current versions of OpenAI tools (now uses DDGS instead of AsyncDDGS)
- Serious error handling fixes and adjustments that are borne not out of Mishal's code, but out of the fact that his bot was created in older versions of Python
- Updated commands to work again using slash commands (discord no longer supports ctx commands like !)
- Added voice function that actually works using ElevenLabs. This also allows you to clone voices and use them to bring your bot to life!