Challenger is a chat game I made where you throw out bold takes — like “the world is flat” — and the AI hits back with witty, sarcastic comebacks. I just pushed the memory update early, so 1.3 is now 1.4 — the AI remembers your last 2 messages for way smoother conversations. You get 3,000 free tokens daily (used per message + reply), and you can buy more with Robux if needed. 1.4 also now has free access (no more 100 Robux to get in). I’m updating faster than I expected and looking for testers to jump in, mess around, and tell me what breaks (or what slaps). There is a system that logs messages (username, user ID, and message) to a private Discord webhook for moderation purposes, but it’s currently turned off until the filtering system works properly — it felt too intrusive for now. The plan is to automatically report users who break Roblox TOS using inappropriate language or behavior. I did try using Roblox’s built-in chat filtering (like TextService:FilterStringAsync), but it doesn’t catch messages outside of their native chat system, so it’s not reliable for third-party AI conversations. Please make sure to read the disclaimer when you join — you can’t play unless you click the X on it. Thanks a ton to anyone testing this out — it really pushed my coding limits. I had to learn how to set up a GitHub repo and deploy a backend with Railway just to get this running. I’m happy to share a copy of the system with the HTTP links and Discord webhook URL edited in the code (so it won’t work unless you set up your own backend and webhook). I’ll be making a forum post or YouTube video soon explaining how to do it — it’s easier than it sounds, but you’ll need at least $5 USD each for both OpenAI API and a Railway subscription. Side note: You can skip Railway by hosting your own backend on your PC, but that’s more work and I haven’t gone down that rabbit hole yet. Railway is cheap and easy if explained right.
Game Link: Challenger 1.4
Clxzed (
