Kailiniai0 % fur · 100 % AI heroes
Menu

Built with AI tools. Sofia still checked the prices. Twice. Out loud.

Guide

An 8-minute YouTube explainer without filming

Ema, a Kailiniai hero

For this guide and its prompts we recommend:

  • Murf records the chapter scripts in one consistent AI voice. Free trial with 10 minutes of voice generation (source)
  • Fliki builds the general scenes from stock footage around the voice track. Free plan (source)
  • PixVerse generates short clips for processes and objects that stock scenes do not cover.

Offers checked 21 Oct 2026 More tools below ↓

You can now make an explainer without filming anything.

One eight-minute explainer with AI voice-over, visuals and a thumbnail, labelled as AI-made.

Adam, a Kailiniai hero

What you need

Sandra, a Kailiniai hero
  • A topic you know well from your own work
  • A search AI that shows its sources
  • Any AI chat assistant, for the outline and script
  • Murf and an AI text-to-video tool
  • Higgsfield and a photo tool for the thumbnail
  • MurfA clear voiceover without booking a studioTrial
  • FlikiA voiced video from text you already haveFree plan
  • PixVerseYou choose from several takes
  • HiggsfieldScenes and motion that look planned

More tools below ↓

Steps

  1. Research with sources and keep them

    Ask a search AI your questions and open every source it gives. Keep a list: claim, source, checked yes or no. Anything without a source you opened stays out of the video.

  2. Write an outline in chapters

    Split eight minutes into a short opening, four or five chapters and an ending. Each chapter answers one question.

    Weak

    Explain everything about home composting.

    Better

    Opening: what goes wrong for most beginners. Chapter 1: what can go in. Chapter 2: where to put the bin. Chapter 3: how to tell it is working. Ending: one thing to do this week.

  3. Record the voice-over in an AI voice

    Write the script chapter by chapter in spoken sentences and record it in one AI voice. Listen to the whole track once and fix words it says wrongly.

  4. Build the visuals from scenes and short clips

    Use stock-style scenes for general shots and short generated clips for processes and objects. No AI children or animals, and no scene that looks like a real person, place or event that did not happen.

  5. Make a thumbnail that matches the video

    Generate a simple image of the topic, add a few words and sharpen it. The thumbnail shows what the video really covers.

  6. Fact-check, label and use the disclosure

    Check every claim against your source list once more. When you upload, say in the description that the voice and visuals are AI-made, and use YouTube's altered or synthetic content setting where it applies.

The full instruction

Copy it into any AI chat assistant and change only the first line.

Change only this line: TOPIC = [how home solar panels work]

You help me make an eight-minute explainer video about my TOPIC for my channel, without filming.

Below this prompt I paste my sources, already checked: each claim with its source. Before you write, ask me one question at a time: who watches and what they already know, and what viewers should be able to do after.

Do four things:
1. An outline: a short opening, 4 or 5 chapters, each answering one question, and an ending with one next step.
2. A spoken script for each chapter, using only the claims in my source list. Where a claim is missing, write [SOURCE NEEDED?].
3. A visual for each chapter: a stock-style scene or a short generated clip. No children, no animals, no real people.
4. A description paragraph that says the voice and visuals are AI-made and lists the sources.

Start with the outline.

Paste your material below this line:

What you end up with

An eight-minute explainer with every claim traced to a source you opened, an AI voice and visuals labelled as AI in the description, and a thumbnail that matches the content.

Common mistakes

  • Keeping a claim because it sounds right, without opening the source.
  • A thumbnail that promises something the video does not show.
  • Uploading without the AI note and the synthetic content setting.

Processes that need more than one tool

Thomas, a Kailiniai hero

Each process below uses several tools, each for the job it does best. Any AI chat assistant works where one is named.

  1. Research and outline

    Tools, in order: a search AI with sources → any AI chat assistant

    1. Ask a search AI your questions, open every source and keep a claim and source list.
    2. Give the checked list to the chat assistant and ask for an outline in chapters.

    The prompt for this process

    Write an outline for an eight-minute YouTube explainer from my checked sources.

Topic: [topic]
Who watches: [audience and what they already know]
Checked claims and sources: [paste: claim, source]

Write a short opening, 4 or 5 chapters and an ending with one next step. For each chapter, give the question it answers, the claims from my list it uses and about how many minutes it takes. Where a chapter needs a claim I do not have, write [SOURCE NEEDED?].

Start with the opening.

  2. Voice-over

    Tools, in order: Murf → Fliki

    1. Record the chapter scripts in one AI voice in Murf.
    2. Import the voice track into Fliki and fix words the voice says wrongly.

    The prompt for this process

    Turn my explainer outline into a spoken script that I will record in an AI voice in Murf.

Outline: [paste]
Checked claims: [paste]
Tone: [e.g. calm, plain, practical]

Write each chapter in short spoken sentences, using only the checked claims. Mark names and terms the voice may say wrongly and suggest a spelling. Add one opening line that says the voice is made with AI.

Start with the opening line.

  3. Visuals

    Tools, in order: Fliki → PixVerse

    1. Build the general scenes from stock footage in Fliki.
    2. Generate short clips for processes and objects in PixVerse and place them in the matching chapters.

    The prompt for this process

    Plan the visuals for my eight-minute explainer: stock scenes in Fliki and short generated clips in PixVerse.

Chapters and their main points: [paste]
Look of my channel: [e.g. clean, bright, close-ups]

For each chapter, list 3 to 5 visuals and say "stock scene" or "generated clip". For generated clips, write a prompt of at most 25 words: one subject, one camera move, the light. No people's faces, no children, no animals, nothing that looks like a real event that did not happen.

Start with chapter 1.

  4. Thumbnail

    Tools, in order: Higgsfield → Fotor

    1. Generate a simple thumbnail image of the topic in Higgsfield.
    2. Correct the light and sharpness in Fotor, add a few words and check it at small size.

    The prompt for this process

    Write a thumbnail plan for my YouTube explainer: image generated in Higgsfield, light and sharpness corrected in Fotor.

Video title: [title]
What the video really covers: [one sentence]

Write 3 image prompts of at most 30 words: one clear object or scene, simple background, strong light. No people's faces, no children, no animals. For each, write thumbnail text of at most 4 words that matches what the video shows. No promises the video does not keep.

Start with prompt 1.

Forward, without brakes

Michael, a Kailiniai hero

Now that the Kailiniai Heroes exist, you can have your own. Create one hero or several, and use them for goals that used to need a whole team or a big budget.

Create your hero

Sofia, a Kailiniai hero
More: Create content with AIAll guides

Create your hero

Join Kailiniai.com, create your own AI hero and reach your goals 5-10 times faster, even the dreams you barely dared to think about.

Create your hero