Summary InfiPlot is a pioneering interactive story game where AI generates all text, images, and audio in real-time, offering unique, personalized narratives without pre-scripted content. It caters to a broad audience seeking tailored fantasy experiences, ensuring endless replayability.

How It Works A multi-agent AI framework (architect, screenwriter, painter, etc.) ensures narrative coherence and consistency. Stories progress through AI-generated scenes, each with a background image and narrative beats. User choices trigger new scenes, often predictively generated for seamless transitions. A vision model interprets background clicks for exploration or advancement. The UI is lightweight HTML, dynamically styled to match the story's aesthetic.

Quick Start & Requirements An online demo is available at infiplot.com. Deployment options include one-click Vercel/Cloudflare (Cloudflare requires a Paid Plan) or Docker for self-hosting. Essential requirements are API keys for AI model providers (OpenAI compatible, Google, Anthropic, Runware, MiMo TTS) and configuring environment variables (e.g., .env.local). Docker setup involves downloading docker-compose.yml and .env.example, then running docker compose up -d.

Highlighted Details

• Fully real-time, AI-generated narrative and visuals.
• Infinite replayability via dynamic, user-driven story generation.
• Multi-provider AI support (text, image, vision, TTS).
• Interactive scene exploration via background clicks.
• Adaptive UI tailored to story aesthetics.

Maintenance & Community Developed by university students, the project is early-stage with a roadmap including mobile apps and ComfyUI integration. Contact: hi@infiplot.com. Community: QQ group 575404333.

Licensing & Compatibility The project links to a LICENSE file, indicating an open-source release. Specific terms require direct review of the license file.

Limitations & Caveats Described as "very early stage," many features are incomplete (e.g., mobile app, save/load, custom characters). Cloudflare deployment needs a paid plan. API costs for AI models are a consideration. Latency, though optimized, may persist.