What happens to your deck, step by step
Five stages between a finished deck and a finished video. You keep control at the script stage, and rendering only starts once you approve.
Every slide becomes a clean page image
Your PPT or PDF is rendered page by page at a fixed resolution, then the text layer is extracted. Charts, tables and diagrams survive because we rasterise the page rather than trying to reconstruct it.
- PPT, PPTX and PDF input up to 40 pages per deck
- Text extracted separately so narration is grounded in the actual slide content
- Pages with no extractable text fall back to visual description
A draft narration you can actually edit
A per-page script is generated from the extracted text and presented for review before anything is rendered. Rewriting a page and regenerating costs nothing until you render, and only edited pages are re-rendered.
- Suggested length per page so narration matches the presenter clip
- Regenerate a single page without touching the rest of the deck
- Warnings when a script is too long for the audio window
Register a presenter from a short clip
A presenter is registered from a short frontal video. Lip movement is driven by the final narration audio, so timing follows your script rather than a pre-baked animation.
- 8–30 seconds of frontal footage, even lighting, no occlusion
- Lip-sync is generated per page from the rendered narration audio
- Unlimited presenters on Team and Enterprise plans
Built-in voices, or clone one you own
Pick from built-in multi-language voices, or clone a voice from reference audio you have the rights to. Cloned voices are gated behind an explicit consent declaration and an approval step, and every clone is logged for audit.
- Multiple languages and speaking styles out of the box
- Voice cloning requires a consent declaration and passes vendor review before use
- On self-hosted deployments you can point narration at a speech model you host yourself
One video, with chapters per slide
Rendered pages are stitched into a single MP4. Each page records its start and end offset, so the finished video can be navigated by slide instead of scrubbed.
- Single MP4 per deck, slide on the left and presenter on the right
- Per-page chapter markers exported alongside the video
- Audio length drives the cut, so narration is never truncated
Run it on a deck you already have
The fastest way to judge output quality is to see your own material rendered. Send one representative deck and we will return a few finished pages.
No credit card. No self-serve signup yet — we onboard pilot teams by hand.