Early access
Explainer videos that show their sources.
Give it a topic or a script. It researches the facts, checks the claims, writes the narration, and builds the video with designed motion graphics. In landscape and vertical. Made for brands and agencies that cannot afford to get it wrong.
We will only email you about early access. Unsubscribe any time.
Made entirely by infovidgen from one short script: "What is a pull request?" Voice, motion graphics, and the live GitHub screenshot are all generated. Press play with sound on.
From idea to finished video
Not a slideshow of stock clips. Each scene is designed for what it says, and every claim is checked before it ships.
Paste a topic or script
Start from a few words and let it research and write, or paste your own narration and add visual hints in brackets.
It researches and checks
A fact sheet with sources goes in. Claims in the narration are checked against it, and anything unsupported is flagged for you.
Edit and export
Preview in the browser, redo a single scene with a note, change a line of narration, then render the MP4. Landscape or vertical.
One script, two formats
The same engine makes a 16:9 explainer for your site and YouTube, and a 9:16 short for Reels, TikTok and Shorts. Layouts are recomposed for vertical, not squeezed.
Landscape, 16:9
Explainers, product updates, and presentations. The video at the top of this page is landscape.
Vertical, 9:16
Large type, word-by-word captions and layouts built for a phone screen. This example explains compound interest in 37 seconds.
Check the numbers
A number on screen counts up to the same figure the narrator says, and claims behind it are checked against the sources.
Vertical, 37 s. Made from a 109-word script.
Built for teams that answer for what they publish
Most AI video tools optimise for speed and volume. We start from accuracy and design.
Sourced, not made up
Narration is written from a fact sheet with sources. Claims are checked, and shaky ones are flagged instead of buried.
Designed, not templated
Dozens of motion-graphic scene types, chosen by what each sentence needs: a map, a number counting up, a timeline, a comparison, a real screenshot.
Editable down to the scene
Change one line, redo one scene, undo a change. You are never stuck with a take-it-or-leave-it render.
Your brand, your assets
Use your own fonts, images and clips, and your own voice by recording over the video. Brand kits are on the way.
Questions
Who is it for?
Marketing teams, agencies, and educators who make explainers, product updates, and short vertical videos, and who need them to be accurate.
How long does a video take?
In our tests, a video under a minute took roughly 10 to 15 minutes to generate end to end, and longer scripts take longer. We are working on making that faster.
What does it cost?
Pricing is not set yet. Join the list and tell us what you would use it for, and that will shape the plans.
Can I use the videos commercially?
We are finalising the licensing details for voices, images and stock footage, and we will publish them before launch. The app already keeps a credits list for any image that requires attribution.
Is the product ready?
It works today and we are making it multi-user. Early access is limited so we can support each team.
Get early access
Tell us where to reach you. We will email once when early access opens.