Type a topic. Get the explainer video.

Or start from your own narration, a PDF, a website or a spreadsheet. KortxAI Motion writes the script, records the voice, animates every scene and renders 16:9 and 9:16. You approve each step before it moves on.

Start a video See how it works Desktop app for Windows
  1. Script
  2. Voice
  3. Scene
  4. Look
  5. Export
  6. Publish
untitled.mp400:00 / 01:00
scene 1

How it works

From topic to published video in six steps

The same steps you see at the top of every project. Each one waits for you, so nothing gets rendered that you have not read and heard.

Step 1 of 6 · Source

Start from anything

A one-line theme, your own narration, a PDF or image, a web page or a spreadsheet. Everything is read on your computer, and you approve what the AI gets to see.

  • PDFs and scanned pages with OCR, with an option to hide personal data
  • Websites become a guided tour of the real page
  • Spreadsheets become charts with your exact numbers
Theme Narration Document Website Spreadsheet
What a context window is and why AI “forgets” the start of a conversation
Step 2 of 6 · Script

Read the script first

Claude, GPT, Gemini or any OpenAI-compatible model drafts about a minute of narration. Edit any line before it moves on, or paste your own text and it is read exactly as written.

Every AI model reads your conversation through a fixed-size window.

When the chat grows past it, the oldest messages fall out first.

That is why a long thread can lose the instructions you gave at the start.

Narrationtarget 240–270 words
Step 3 of 6 · Voice

Give it a voice

Pick a narrator from Deepgram, OpenAI, ElevenLabs or Microsoft Azure, Brazilian Portuguese included, or use your own recording. Every word is timed for the captions.

  • Hear a sample of each voice before you choose
  • Fix one sentence without redoing the video
Narration
Voice
Francisca · pt-BRAntonio · pt-BRAva · en-US
0:12When the chat grows past it, the oldest messages fall out first.Re-record
Step 4 of 6 · Scene

Scenes that explain

The AI animates the whole narration, timed beat by beat: a single card that changes shape, a zoom map that travels from idea to idea, or a tour of a real website.

  • Every frame is measured and layout problems are fixed without another AI call
  • Ask for a change in plain words, or edit on the canvas
old messages slide out
layout checkedversions savedcanvas editor
Step 5 of 6 · Look

Make it yours

Switch the theme, the drawing style and the captions at any time. The scene is rebuilt on the spot, with no AI cost.

  • Flat, Sketch or Cartoon, on six themes including your own colours
  • Karaoke, box, bouncing or plain captions
  • Ten presenter mascots with lip sync, or your photo
  • Sound effects and music that ducks under the voice
Style
Flat
Sketch
Cartoon
Theme
CasaNoitePapel TechPopCustom
the oldest messages fall out
Step 6 of 6 · Export and publish

Render, then publish

Native 16:9 and 9:16, with or without your own video beside the animation, plus the subtitle file. The publishing kit is written for you.

  • Covers in both formats, three title options, a description with chapters and tags
  • Your logo, watermark and an end card with your handle
16:9
9:16
MP4 16:9MP4 9:16With your videoSubtitles .srtCoversTitles and chapters

Real data

Your numbers, never retyped

Drop in an Excel file or a CSV. Motion reads it on your computer, suggests the charts and shows you every value before anything goes to the AI.

The AI decides where each chart appears and when. The figures themselves come straight from your sheet, with the currency from the cell format and numbers and dates written the way the video’s language writes them.

ProductRevenue
BackpackR$ 21.980,75
NotebookR$ 12.450,50
PlannerR$ 9.870,00
Ranked from largest to smallestIllustrative data

Languages

One animation, many languages

Make a version in another language and Motion translates the script, records a new voice, redoes the captions and translates the text on screen. The animation is reused and re-timed sentence by sentence to the new narration.

The app itself speaks Portuguese, English, Spanish, Italian, French and German.

ENEnglish · original
PTPortuguês (Brasil)
ESEspañol
DEDeutsch
FRFrançais
ITItaliano
NLNederlands
JA日本語

What you get

One topic, one folder ready to publish

Every video is a folder on your disk with everything you need to post it. Nothing is locked in a cloud.

  • context-window_16x9.mp4YouTube, LinkedIn
  • context-window_9x16.mp4Shorts, Reels, TikTok
  • context-window_16x9_video.mp4with you beside it
  • legendas.srtsubtitles
  • capa_16x9.jpg · capa_9x16.jpgcovers
  • publicacao.mdtitles, description, chapters, tags
  • fonte/script, voice, scene and versions

Batch production

A whole series in one go

Paste one topic per line, or a title and a short brief. Motion runs each one from script to finished video while you work on something else.

Organise everything in project folders, one per client, series or campaign, and see what each video cost in AI usage.

  • What a context window is ready
  • How AI agents screen résumés ready
  • Buy now, pay later in a minute rendering
  • What is a vector database voice
  • When to use RAG waiting

Your keys, your files

You stay in control of the AI and the cost

Bring your own keys

Pay the AI providers directly

Connect the services you already use: Anthropic, OpenAI, Google Gemini, OpenRouter or a local model, plus your voice provider. Keys stay in the Windows Credential Manager, never in a file.

Local first

Sources read on your computer

PDFs, images, websites and spreadsheets are read locally. You review the text, and only what you approve goes to the AI, with the estimated cost shown first.

Cost per call

No surprise at the end of the month

Every AI call is logged with its real cost per project and per step, with a monthly total, an editable price table and a CSV export.

Plain folders

Your videos are your files

Group videos in project folders and move them whenever you like. Everything stays on your disk.

FAQ

Questions people ask

Do I need my own API keys?

Yes. One AI key for the script and the scenes (Anthropic, OpenAI, Gemini or an OpenAI-compatible service) and a Deepgram key for the captions, plus the voice provider you choose. The app has a Test button for each key.

Is there a Brazilian Portuguese voice?

Yes. Microsoft’s neural voices through Azure (Francisca, Antonio, Thalita and others), plus OpenAI and ElevenLabs, which also speak Portuguese. Or use your own recording.

Can I use my own voice or video?

Yes. Import an audio file or a video of yourself speaking. The app transcribes it word by word, animates the scene over your speech and, for a video, also renders the formats with you beside the animation.

Does the AI make up numbers?

It is told not to. Without real data, a chart is labelled as illustrative. With a spreadsheet, the figures come straight from it and the AI retypes none of them.

How much does a video cost to make?

It depends on the model and the voice you pick, and you pay each provider directly. The app shows an estimate before each step and the real cost afterwards, per project and per month.

Does it run on a Mac?

For now it is a desktop app for Windows 10 and 11. It uses the installed Edge or Chrome to build and record the scenes.

Your next video starts with one sentence.

$12 a month. Try it on a topic you explain every week.

Windows 10 or 11 · Microsoft Edge or Chrome installed · your own AI and voice keys

Start a video