What it actually costs to convert a deck to video in 2026
Per-slide, per-minute and per-seat pricing compared, what the headline number leaves out, and where the real cost sits for a team maintaining a library rather than making one video.
Pricing for this category is deliberately hard to compare, because three different units are in use and the headline number is rarely the cost you end up paying.
The three pricing models
Per minute of finished video. The large avatar platforms price this way. It is intuitive and punishes long decks, which is where training content lives.
Per seat, per month. Most general video tools. Fine for a marketing team, awkward for training, where the people who need to watch vastly outnumber the people who need to make, and charging per viewer is the wrong axis.
Per slide. Cost tracks pages, which is what actually drives the compute: rendering, text-to-speech and lip-sync all scale with how many slides there are.
Per-slide has one property the others lack: you know the number before you start. The page count is read from your file at upload, so the price is quoted before anything runs. With per-minute pricing you only know the cost once the video is finished, because the audio length is not known in advance.
Some actual numbers
DocPresenter charges 10 credits per slide, with a two-slide minimum and a 40-slide ceiling. A 20-page deck is 200 credits, which works out at roughly:
- $3.90 on the smallest one-off credit pack
- $2.45 on a monthly Lite subscription
- $1.44 on the yearly plan
Those 200 credits buy about 40 minutes of finished video, so the effective rate is roughly $0.04 to $0.10 per finished minute, depending on which tier you are on. Signing up includes 60 credits, which covers a six-page deck — enough to convert a real extract rather than a demo file.
For scale, the large avatar platforms sit around $1.27 and $0.78 per finished minute. The gap is roughly 8–29×, and it comes from where the work runs: a single-purpose GPU pipeline rather than a general-purpose creative platform with an editor, a stock library and a collaboration layer attached.
What the headline number leaves out
Re-recording, if you record. The sticker price of a screen recorder is zero and its real cost is a person’s afternoon every time the deck changes. For anything with a review cycle this dwarfs the conversion fee. It is the argument in four ways to add narration.
Editing time. Whichever tool you use, someone has to review the script. Budget minutes per deck, not seconds per slide. This is the one cost that does not shrink with a cheaper tool.
Rendering wait. Not money, but it is a real cost if you have deadlines. Conversion is GPU work and does not parallelise freely — submit long decks the day before you need them.
Other languages. With recorded narration each language is a full production. With generated narration it is a translation and a re-render, so the marginal cost of a fifth language is the translation itself. See localising a deck.
What is not included. No stock footage library, no timeline editor, no interactive branching. If you need those, they are different products with different prices, and adding up features you will not use is not a saving.
How to work out your own number
- Count your slides, not your videos. A 20-slide deck is 200 credits regardless of how long the narration runs.
- Multiply by how often you will re-render. A quarterly-updated deck converts four times a year; that is the number to budget, not the one-off.
- Add script review time. This is usually larger than the conversion cost for a small library.
- Compare against the alternative’s update cost, not its sticker price.
The last step is the one that changes the answer. For a single video that will never change, most tools are cheap enough that it does not matter. For a library that gets updated, the per-update cost is the whole decision.