Skip to main content
Upload a PDF deck and Origami AI turns each slide into a fully narrated video — no cloud rendering, no subscriptions. The AI drafts natural speech for every slide, Kokoro.js synthesizes the audio directly in your browser, and FFmpeg.wasm composes everything into a polished 720p or 1080p MP4 you can download and share.

​
How It Works

1

Upload your PDF

Drag your presentation file onto the upload area or click to browse. Origami AI extracts slide images and reads the text from every page automatically — no manual copy-pasting needed.
2

Generate narration scripts

Click Generate Narration and the AI drafts a natural, conversational script for each slide. You can use a local model (WebLLM, runs entirely on your device) or a cloud API you’ve connected in ⚙️ Settings → API — whichever you prefer.
3

Synthesize voice audio

Origami AI uses Kokoro.js to convert every script into high-quality speech audio — no server round-trip, no file uploads. Choose from multiple voices in Settings → General or directly in the Slide Editor.
4

Edit in the visual editor

Fine-tune anything before you export: rewrite individual scripts, adjust audio timing and post-slide delays, pick transition effects between slides, set your preferred aspect ratio, and add background music with automatic ducking under the narration.
5

Export the video

Click Render Video to compose all your slides, audio tracks, music, and transitions into a finished MP4 — entirely in your browser. Choose 720p for faster rendering or 1080p for the highest quality.

​
Key Controls

The Slide Editor gives you per-slide control over every aspect of your video:

​
Aspect Ratios

Origami AI supports four aspect ratios so you can match your output to any platform or screen:

16:9 — Widescreen

The standard for presentations, YouTube, and desktop viewing. Best choice for most PDF decks originally designed in landscape.

9:16 — Vertical

Portrait format optimised for YouTube Shorts, Instagram Reels, and TikTok. Great for repurposing slide content as social video.

4:3 — Classic

Traditional presentation format. Useful when your slides were designed for an older 4:3 projector ratio.

1:1 — Square

Equal width and height for Instagram feed posts and multi-platform sharing.
Set your preferred aspect ratio in ⚙️ Settings → General before rendering.

​
Export Options

When you’re ready to export, Origami AI gives you two resolution choices:

720p MP4

Faster to render and smaller in file size. A great choice for drafts, social sharing, or machines with limited resources.

1080p MP4

Full HD output for polished presentations, uploads to YouTube or LinkedIn, and anywhere quality matters most.
You can also export your entire project as a .origami archive (File → Export Project). This portable ZIP contains all slides, audio blobs, scripts, and settings — import it on any device running Origami AI to pick up exactly where you left off.
On a slower machine, switch the TTS quantization to q4 in ⚙️ Settings → TTS Model. The q4 model processes audio significantly faster with only a small reduction in voice quality — ideal for drafting or long decks. Switch to q8 when you need the highest audio fidelity for your final render.
The first time you use a feature that requires an AI model (narration generation or TTS), Origami AI downloads the required model to your browser’s local cache. This one-time download can range from a few hundred MB to a couple of GB depending on the model. After that, everything runs offline with no further downloads needed.