How to Create Educational Videos with AI: ChatGPT, Sora, and CapCut

---

Artificial intelligence is transforming the way we create content. In this tutorial, we guide you through the process of creating a comprehensive educational video—research, script, visuals, voice, and avatars—using three free or low-cost AI tools: ChatGPT, Sora, and CapCut.

Whether you're an educator, content creator, or student, this guide will help you produce engaging video lessons without needing to film yourself or write code.

Step 1: Build the Knowledge Base with ChatGPT

The first step in any video is gathering accurate, up-to-date information. Instead of researching manually, we use ChatGPT’s Deep Search to pull the most recent insights from across the internet.

Example of a ChatGPT deep search request focused on AI in data science (July 2025)

We provide a topic and a few parameters:

• Target audience: General public

• Level: Introductory

• Focus areas: AI applications in data science, social media, GIS

• Time frame: Past 1 month (AI evolves fast)

ChatGPT scans 20+ sources and produces a report. This becomes our knowledge base for the next step.

Step 2: Use ChatGPT to Create Educational Videos with AI

With the content ready, we now ask ChatGPT to write a 3–5 minute script for our video. We also ask it to generate image prompts for each section of the video.

ChatGPT produces a full script and a list of image prompts to visualize each segment

The result includes:

• A complete script with intro, 2–3 content sections, and a conclusion

• Prompts like “AI in Data Science – July 2025” or “Generative Maps using OpenAI”

You can paste this script directly into a video editor, or generate visuals with the next tool.

Step 3: Create Visuals with Sora

We utilize Sora, an AI image generation tool, to transform each prompt into a relevant and high-quality image.

 Sora generates high-resolution visuals from the prompts provided by ChatGPT

You can choose:

• Format: Vertical, horizontal, or square (depending on your platform)

• Number of versions: Usually, one high-quality image per scene is enough

These visuals serve as key frames in the final video.

Step 4: Assemble the Video in CapCut

Now, we open CapCut, which features an Instant AI Video option. It lets you paste your script and automatically creates a video with voice narration and visuals.

CapCut’s AI video builder uses your script and generates a full video with voice, transitions, and styles

Steps:

1. Paste your script

2. Choose a voiceover (you can also upload your voice)

3. Select the style and aspect ratio

4. Choose whether to add an AI avatar

CapCut automatically syncs visuals and voice.

Optional: Add Avatars

To make your video feel more engaging, CapCut lets you add an AI avatar presenter.

Select an AI avatar to act as a virtual presenter—no filming required

You can:

• Choose from male or female avatars

• Decide which segments use avatars and which display images

• Replace stock visuals with Sora-generated visuals for better accuracy

  
  

Final Touches: Replace Images, Add Captions, and Polish

Once the video is generated, you can:

• Replace CapCut’s default images with the Sora visuals

• Decide which segments show avatars and which show visuals

• Add captions and optional background music

The final result is a professional, AI-generated educational video created in under an hour.

Watch the Full Process

You can watch the full step-by-step video here:

Why This Workflow Works

This method allows anyone—teacher, student, content creator—to:

• Automate research with ChatGPT

• Create visuals with Sora

• Produce and narrate videos in CapCutAll without coding or recording on camera.

It’s fast, accessible, and scalable—perfect for modern learning environments.

Bonus Tips

• Use vertical format for TikTok or Shorts

• Use horizontal format for YouTube

• Add chapter markers in your video to help viewers skip to key sections

• Replace avatar voices with your own for a more personal touch