LLM Noire

LLM Noire

AI-Native Mystery Game

4 followers

LLM Noire is an AI-Native mystery where players read a description about a case, evidence, and three suspects all AI generated. The player then enters room A, B, C to interview said suspects using natural language and ultimately deduce who committed the crime.
LLM Noire gallery image
Free
Launch tags:Free Games•Artificial Intelligence•Games
Launch Team
Checkmarx Developer Assist for AI IDEs
Security linter for vibe coding: fix vulns as you build
Promoted

What do you think? …

Chris Pravata
Maker
📌
About LLM Noire LLM Noire is a mystery game I built to break away from the limitations of traditional detective games. Most mystery games follow a fixed storyline—with scripted dialogue, linear puzzles, and one correct solution. Once you’ve played through, the experience is over. I wanted to change that. With LLM Noire, every case is different. Every suspect has a backstory that changes. Every clue can be interpreted multiple ways. Powered by large language models, the game lets you interrogate suspects, analyze evidence, and form your own theories—using natural language, not preset dialogue trees. There’s no single way to solve a case, and no two playthroughs are the same. I built LLM Noire because I believe mystery games should reward curiosity, creativity, and critical thinking—not just remembering patterns. It’s an open-ended sandbox of crime and deduction, designed to feel more like a real investigation than a scripted experience. What makes it special is the freedom: you’re not guessing what the game wants you to say—you’re deciding what you think is true. And the suspects? They push back. They lie, deflect, and evolve. That’s what makes each case feel alive. This isn’t just a murder mystery. It’s a new kind of storytelling engine—one that reacts to you, not the other way around.