Introduction
Welcome to MovingLines documentation
Welcome to MovingLines
MovingLines is an advanced AI orchestration platform designed to generate mathematically beautiful, story-driven, and cinematic animations using Manim Community Edition.
By leveraging Gemini 2.0 Flash and a custom Agentic Workflow, MovingLines moves beyond simple code generation to become a true Narrative Designer.
The "Aha!" Engine
The core of MovingLines is the "Aha!" Engine — a multi-agent system that prioritizes educational clarity and visual continuity.
- Phased Generation: We don't just write code; we storyboard first.
- Progressive Revelation: Content builds step-by-step to maximize understanding.
- Visual Continuity: Smooth transitions and morphs instead of static cuts.
Key Components
Explore our deep-dives into the system's architecture:
Architecture
A high-level view of how the Frontend, Backend, AI Engine, and Renderer communicate. Learn more
The Narrative Planner
Learn how we transform a single sentence prompt into a detailed cinematic storyboard. Learn more
RAG & Memory
See how we use Pinecone to provide the AI with high-quality, stable Manim code examples. Learn more
Anti-Crash Sanitizer
Our specialized engine that detects and fixes LLM hallucinations before they reach the renderer. Learn more
Quick Start
Ready to dive in? Head over to Getting Started to set up your local environment or deploy via Docker.
Quality & Performance
MovingLines supports a variety of rendering options to fit your workflow:
- 480p (15fps): Instant previews for rapid iteration.
- 720p/1080p: Standard sharing and HD output.
- 4K (60fps): Premium cinematic quality for final exports.
Made for the Math & Motion Community