Discord bot hosting
that never goes offline.

A Discord bot is a small program that holds an open connection to Discord and reacts to what happens in your servers. It only works while that program is running. On your own machine that means the bot dies when you close the laptop, when Windows updates overnight, or when your internet drops for thirty seconds — and your members notice before you do.

Discord bot hosting is simply that program running somewhere that never turns off. You upload your code, pick the language it is written in, press start, and the bot stays online whether or not your computer is. If the process crashes, the panel restarts it. If you push a fix to Git, you pull it in from the file manager and restart.

Setup is instant: pay, and within a minute the panel login is in your inbox with an empty server waiting for your files. If you want the whole thing step by step first, read how to host a Discord bot around the clock.

Pick the runtime that matches your code when you order; it takes one click and can be changed later from the panel. Node.js runs anything built on discord.js, and also runs TypeScript directly without a build step. Python covers discord.py and Pycord, which is what most tutorial bots are written in. Java runs JDA and Discord4J projects from a jar.

Version choice is yours from the panel — recent Node.js and Python releases are available, and switching is a dropdown and a restart, not a support ticket.

If you are not sure which you need, look at your project: a package.json means Node.js, a requirements.txt means Python, a .jar means Java.

Less than most people expect, and more than the free tiers give you. A bot is not compute-heavy — it waits on a websocket and reacts — so the memory it uses is mostly the library and the cache of servers and members it keeps in memory.

In practice: a bot written in Go idles around 15 to 30 MB, one in Python around 50 to 80 MB, one in discord.js around 60 to 120 MB and noticeably more once it has cached a few large servers, and a Java bot on JDA starts around 150 to 250 MB before the JVM overhead. That is why our entry plan is 512 MB — comfortable for a real bot on any of those, where the common free tier gives 256 MB.

Move up when your bot joins more servers, caches more, handles media, or when you want to run several bots side by side. Upgrades keep your files, your database and your address.

Music bots are the one Discord workload that genuinely uses CPU, because audio has to be decoded and re-encoded in real time for every listener. That is what Lavalink exists for: it moves the audio work out of your bot into a dedicated Java service, so your bot stays responsive while music plays.

You can run Lavalink alongside your bot on the same plan if you have the RAM for both — the 2 GB tier upward is a sensible starting point — or keep it on its own server so a busy queue never slows your commands down.

If you would rather not write a music bot at all, ask us: we can set up a ready-made one for you on your plan.

Every bot here runs in its own isolated container on Pterodactyl, the same open-source panel used across the game hosting industry. If you have run a Minecraft server in the last few years, you already know the interface.

Practically, that means a live console you can type into, a file manager that handles drag-and-drop and archives, sFTP with your own credentials, scheduled tasks, resource graphs showing real CPU and memory use, and subusers so a co-developer can help without getting your billing account.

Isolation matters more than it sounds: your bot has its own memory and its own disk, so a neighbour with a memory leak is not your problem.

Most bots that do anything interesting need to remember something — levels, warnings, economy balances, per-server settings. Every plan includes at least one MySQL database with its own credentials, reachable from your bot, so you are not stuck writing to a JSON file that a restart can corrupt.

Storage is NVMe throughout, from 8 GB on the entry plan to 100 GB on the largest. That is generous for code and logs, and leaves room for cached media if your bot handles it.

Automatic backups are included, and you can trigger one by hand before a risky deploy. Restoring is one click.

Our nodes are in Europe, behind Voxility filtering with over 1 Tbps of capacity across 23 points of presence. For a Discord bot, latency to your users barely matters — your bot talks to Discord's servers, not to your members directly — so what actually counts is that the connection to Discord is stable and the machine is not oversold.

We back it with a 99.9% Uptime SLA and we do not oversell memory. If a node needs maintenance you get told in advance, and your bot comes back on its own afterwards.