The Universal Interactive Engine
Upload a PDF, scan a repository, or paste raw text. Origami breaks it down, streams an interactive explanation, and generates a **v0-powered MVP route** in seconds.
Explore Workspaces
Discover interactive canvases generated automatically from source documents. Click to experience real-time AI transformations.
3
Source formats
10s
Avg workspace load
5+
Interactive components
∞
Documents to fold
Scan any repo.
Understand it instantly.
Paste a public GitHub URL and Origami crawls every README, CONTRIBUTING guide, and package manifest across the entire default branch. No cloning, no setup — just instant understanding.
- Automatic discovery of markdown & manifest files
- Per-file AI analysis with cross-reference awareness
- Architecture graph generated from real repo structure
- Dependency dashboard from package.json / pyproject.toml
Canvas Capabilities
Stream live, interactive
AI components.
The interactive canvas isn't just a static summary. Origami streams a real-time, AI-selected component — architecture graph, scenario simulator, or package dashboard — purpose-built for your exact source type.
- Force-directed architecture graph for repos
- Real-time streaming with instant rendering
- Full-screen expandable modals for deep dives
- Automatically re-analyzes on source mutation
From source to app.
Ready to ship.
The **v0 MVP Builder** reads your source, understands the domain, and produces a fully self-contained, responsive single-page web app — styled, interactive, and powered by the v0 API.
- Complete HTML/CSS/JS app with zero external config
- Domain-aware design adapts to your specific content
- Persisted in-browser with a shareable /mvp/[id] link
- Copy entire source as a standalone file in one click
What drives Origami
Speed above all
From source to interactive workspace in under 10 seconds. We obsess over latency at every step of the pipeline.
Universal intake
Any document format, any domain. If it contains information, Origami can fold it into something useful.
Developer-first
Built by developers, for developers. Every API is clean, every component is composable, every output is copyable.
Open by default
We believe AI tooling should be transparent. The source is readable, the outputs are auditable, the logic is explainable.
How we got here
Concept
Initial prototype exploring AI-powered document analysis with a basic chat interface.
Source engine
Built the universal source intake system supporting PDF, text, and GitHub repos.
Interactive canvas
Introduced the streaming architecture graph and interactive breakdown components.
v0 MVP generator
Launched one-click single-page app generation from any source document.
Workspace v2
Redesigned the workspace with tabbed sidebar, drawer UI, and Q&A functionality.
Built with the best
A modern, edge-ready stack optimized for speed, reliability, and real-time generation.
v0
FrontendFront end developed by v0 for rapid prototyping and high-fidelity design
Vercel AI SDK
AIStreaming AI responses, tool calling, and useChat hook
OpenAI GPT-5.5
AIState-of-the-art reasoning model powering all analysis and generation
React Flow
VisualizationInteractive node-graph rendering for architecture diagrams
Tailwind CSS
StylingUtility-first styling with a curated dark design system
TypeScript
LanguageEnd-to-end type safety across the full stack
The Origami identity
Origami
The word "origami" means folding paper — transforming something flat into something dimensional and meaningful. We do the same for your data.
#CCFF00
Lime 300 — chosen for its vibrancy, representing energy, clarity, and rapid digital growth against our pure dark mode canvas.
Ready to fold?
Upload a PDF, paste a GitHub link, or drop in raw text. Your interactive workspace is generated in seconds.
Launch Workspace