# AI documentary maker > Write what the video is about, or start from a transcript made here. You get a script of scenes to edit, free; then a voice reads it over a picture for each scene, with slow camera moves and captions, as an MP4. Source: https://clipscript.app/ai-documentary-maker · Updated: 2026-10-31 · Publisher: ClipScript ## Make a documentary-style video Paste a topic, notes, an article or a transcript; get an editable script of scenes, then a narrated video with a picture per scene and captions, priced per minute. ## How does this AI documentary maker work? A language model turns your topic, notes, article or a transcript from this site into a script of scenes: a few sentences of narration and one picture each. The script is free and every line is editable. Press Make and a voice reads each scene over a picture an image model paints, moving slowly while the voice speaks; the scenes are joined with captions into an MP4. 40 credits per started minute of narration, worked out from the script first. - **Script:** Free, editable, any language - **Video:** 40 credits a minute, up to 10 minutes - **Look:** Documentary photo, illustration, cinematic, watercolour - **Out:** MP4, 1920×1080 or 1080×1920 ## From idea to video 1. **Give it the material** A topic, notes, an article, or a transcript from the YouTube or upload tools here. Pick a length and the narration language. 2. **Edit the script** Every scene shows what the narrator says and what the picture should show. Rewrite lines, fix facts, cut scenes, add your own. The price updates with the word count. 3. **Make it** Choose the picture style, the voice, 16:9 or 9:16 and captions. The render takes a few minutes; download the MP4 within 48 hours. ## What to know before you make one - **Pictures, not footage** Each scene is one still picture made by an image model, with a slow zoom. It suits explainers, history, essays and storytelling; it is not stock footage or AI video clips. - **Check every fact** The script is written only from what you give it and told not to invent; models still slip. Read it before you render. The voice reads exactly what the script says. - **Your channel, your rules** YouTube does not pay for mass-produced or repetitive content (its July 2025 "inauthentic content" update). Use this for videos you write and stand behind, not for templated uploads in bulk. - **Refunded if it fails** The price is charged when the render starts and returned automatically if it fails. A video that renders but that you would change is not refunded; edit the script first. ## AI documentary questions ### What does a video cost? 40 credits per started minute of narration, from the script at 150 words a minute: a 3-minute script is 120 credits, a 10-minute one 400. 1,500 credits are $15. Writing and rewriting the script is free. ### Can I make faceless YouTube videos with it? Yes: a voice over pictures, no camera. Write or edit the script yourself; YouTube's monetisation rules exclude mass-produced, repetitive videos, and a generic script read over generic pictures is exactly that. ### Which voices are there? Six narrator voices, named alloy, echo, fable, onyx, nova and shimmer, three lower and three higher. They read many languages; we have not rated their accents. ### How long can a video be? Up to 10 minutes of narration, 60 scenes. For a longer film, make it in parts and join them in your editor. ### Can I start from a YouTube video or a podcast? Yes: get its transcript on the [YouTube transcript](https://clipscript.app/youtube-transcript) or [audio](https://clipscript.app/audio-to-text) page, then use "make a video from it". Rewrite it in your own words before you publish someone else's ideas. ### How long does it take? A few minutes for a short video; it grows with the number of scenes. We have not timed production renders yet and will publish the figure when we have. ## Make your first video Write a topic, edit the free script, pay only for the minutes you render. [Start](https://clipscript.app/)