So I built this little thing called an AI narrative story maker and I want to be honest about why.
I grew up on fantasy books, manga, manhwa, anime, JRPGs, MMOs. Like every kid who reads too much, I spent half my brain wondering what I would do if I were in there. What would I do at that tavern? Which side would I pick? Would I trust the cursed knight? Childish, sure, but if you're reading this you've probably done the same.
Then you grow up, you stop. You read a manhwa here and there. Life happens.
Recently I finally managed to leave my 9-5 and suddenly I had time again. As a long-time fan of Omniscient Reader's Viewpoint (if you know, you know), I had this stupid idea. What if I built a kind of story-RPG where the AI is the narrator and you're the character who decides what happens next. I know plenty of apps already do this. But I always wanted to make a game, always wanted to write a fantasy book, and I never did either. So this is more about doing the thing than about market logic.
The app is called First Person Viewpoint (FPV). It's on iOS and Android. You make up your own world, create your character, and the AI handles the narration while you take actions. Basically the same family as AI Dungeon and the others, but vertical-scroll, no chat bubbles, reads more like a manhwa than a chat window.
Genres are pretty open: fantasy, sci-fi, romance, manga, horror, custom, even NSFW (the app is 18+ only). There's a small community feed where users can publish their own worlds.
Honest about the business side
- We don't own an LLM. We're not OpenAI. We use third-party APIs and somebody has to pay the per-token bill.
- Free tier has ads, just enough to cover token cost on free users so I don't have to sell a kidney every month.
- Paid tiers are ad-free with more actions and a better narration model.
- We don't sell your data. Sessions live mostly on your device, the server only has what it needs.
It's pretty fair I think. I'm not trying to flip it for a billion. I just wanted it to exist.
What I'd actually love feedback on
Right now most users are friends and family. If you have iOS or Android and you want to try it, I'd be very grateful for any honest reaction. If something feels broken or confusing, even more grateful, that's the feedback that actually helps a solo dev.
The things that keep me up at night:
- The first 60 seconds of onboarding. Does the "you are the character" framing land or feel like work?
- The "older text fades to navy" effect. Cool or annoying?
- Free tier at 250 actions per month. Too tight or about right?
- Pricing. Premium is $9.99 and Max is $24.99. Does that read as fair or as another AI tax?
Where this kind of stack is wrong
I'm not going to pretend this approach fits every project. If you're considering shipping a solo AI app, here is when I'd say do not bother:
- You need offline first. LLM-driven apps need a network round trip per action. Not viable on a plane.
- You need real-time voice. Text-stream latency is fine for prose, terrible for live conversation.
- You can't tolerate per-token costs that scale with usage. Pick a flat-fee API or run your own model on a fixed-cost server. The unit economics of a free tier paid for by ads are tight.
For everyone else, the maths is doable. The pieces exist. The hard part is the taste decisions, not the wiring.
So
I built the game I always wished existed. It is live. People are playing it. The kid in me who used to sit in the tavern wondering which faction to join finally got to actually decide.
If you have shipped a solo project that started as a "stupid idea you couldn't shake", I want to hear it. What is the dumb thing you finally got around to making?
Top comments (0)