Quick Start
Get up and running with Origami AI in minutes — no installation required.
PDF to Video
Upload a presentation and export a polished narrated MP4 video.
Shorts Generator
Turn any topic into a vertical short with AI visuals, voiceover, and captions.
Screen Recording
Record your screen with cinematic auto-zoom driven by your cursor activity.
What You Can Do with Origami AI
Origami AI combines six powerful tools in a single browser tab:PDF → Video
AI narration scripts, TTS audio, and full MP4 export from your slide deck.
Shorts Generator
AI script, Pollinations visuals, Kokoro TTS, and burned-in captions.
Screen Recording
Smart auto-zoom with optional Chrome extension for richer telemetry.
AI Assistant
Chat with local WebLLM models — attach images and videos for analysis.
Voice Studio
Generate high-quality TTS audio from any text using Kokoro voices.
File Studio
Convert and compress images and audio files entirely in your browser.
How It Works
Origami AI uses WebGPU to run AI models directly on your device’s GPU. Scripts are generated by local WebLLM models (or your configured cloud API), speech is synthesized by Kokoro.js, and video is composed by FFmpeg.wasm — all without leaving your browser tab.1
Open the app
Visit origami.techmitten.com — no sign-up required to start.
2
Choose your tool
Pick PDF to Video, Shorts Generator, Screen Recording, AI Assistant, Voice Studio, or File Studio from the navigation menu.
3
Download AI models (first time only)
On first use, Origami AI downloads a lightweight AI model to your browser cache. This takes a few minutes but only happens once.
4
Create and export
Generate, preview, and export your content directly to your device — nothing is uploaded.
Origami AI works best in Chrome or Edge 113+ where WebGPU is fully supported. See Requirements for browser compatibility details.
Get Started
System Requirements
Check browser and hardware requirements before you begin.
Configuration
Customize AI models, voices, and API connections.
First Video Guide
Step-by-step walkthrough of the full PDF-to-video workflow.
Troubleshooting
Solutions for WebGPU, rendering, and TTS issues.