How to Create AI Shorts in Under 10 Minutes: A Step-by-Step Workflow

Gone are the days where creating a single shorts video took hours of preparation followed by the hassles of actually shooting it. With the proper workflow using AI, you can now generate an entire short video in less than 10 minutes! 

This guide provides a detailed, step-by-step workflow to do just that, using a combination of powerful AI tools. By leveraging platforms like ChatGPT for scripting, Leonardo AI for visuals, and Revid AI for video assembly, creators can drastically speed up their content production. So whether you're a seasoned creator or just starting, this workflow can potentially change how you make content.

Introduction

Creating engaging short-form video content can be time-consuming, but AI is changing the game. This guide demonstrates a practical, end-to-end workflow for producing a complete short video from scratch in under 10 minutes.

You’ll learn:

• The exact tools used for scripting, image generation, and video assembly.
• A full, three-step workflow from concept to final video.
• Pro tips and best practices that ensure high-quality results.
• Observations from real tests 

What this guide covers

• A complete workflow to create AI shorts quickly and efficiently.
• The specific tools and their roles in the creation process.
• Insights from real testing and tool comparisons.
• Tips to get the best results and avoid common pitfalls.

Tools & Their Roles

To create our AI short, we used three primary tools, each with a specific function:

1. ChatGPT: Script & Prompt Generation
   ChatGPT was used for two key tasks: first, to generate the initial video script on the topic "Why we procrastinate," and second, to convert the finalized script lines into image prompts for our visual generator.
2. Leonardo AI: Visual Generation
   For creating the images, we used Leonardo AI. It’s excellent for generating high-quality visuals from text prompts. We specifically used the Fluxscale model, which offers a great balance between quality and cost, consuming fewer credits than other high-end models.
3. Revid AI: Video Assembly & Narration
   Revid AI was the final step, stitching everything together. We uploaded our generated images, added a voice-over and background music, and let the tool assemble the final short video. Its strength lies in turning a sequence of static images into an engaging, narrated video.

The 3-Step Creation Process

This workflow is broken down into three simple stages, moving from text to visuals and finally to the finished video.

Step 1: The Script (ChatGPT)

The foundation of any good video is the script. We started by asking ChatGPT to generate a short, punchy script that could be narrated in about 20-25 seconds.

📌 Prompt:

Create a short script for a 20-second video on the topic "Why we procrastinate."

📌 Output (Final Script Used):

Why do we procrastinate? It's not because we're lazy. It's because our brain is trying to avoid discomfort.

When a task feels boring, stressful, or overwhelming, our brain pushes us toward quick rewards instead—like scrolling or watching Netflix.

The way out? Start small. Even the tiniest step reduces anxiety and builds momentum.

Procrastination isn't a time problem—it's an emotional problem. Manage that, and the work gets done.

Step 2: The Images (ChatGPT + Leonardo AI)

With the script finalized, the next step was to create visuals for each line. We fed the script back into ChatGPT to generate detailed image prompts.

📌 Prompt:

Turn each line of the following script into a detailed image prompt for an AI art generator.

📌 Output (Image Prompt Examples):

This produced a unique prompt for each key phrase. Here are a few relevant examples from the sequence to show the narrative progression:

• To set the scene:
  "Why do we procrastinate?"
  → Minimalist illustration of a person sitting at a desk, staring at a blank page, looking puzzled.
• To show the conflict:
  "our brain pushes us toward quick rewards instead..."
  → Illustration of the same person reaching for their phone with a little sparkle above it.
• To show the resolution:
  "Manage that, and the work gets done."
  → Illustration of the person happily finishing their work at the desk, with a paper or screen showing a checkmark.

Step 3: The Video (Revid AI)

This final step doesn't require a text prompt but involves a series of actions within the Revid AI platform to assemble the video.

📌 Process:

1. Upload Images: We uploaded all 11 of the generated images from Step 2 into Revid AI, ensuring they were in the correct narrative order.
2. Select Media Options: Chose the "use your own media" setting.
3. Add Audio: Selected a suitable narration voice and background music from the available libraries.
4. Generate: Clicked the "Generate" button to compile the video.

Step-by-Step Workflow Summary

Follow this exact order to create your own AI-generated short:

1. Generate & Refine Script in ChatGPT:
2. • Start with a topic and ask for a short script.
   • Edit the script down to 20-25 seconds for optimal pacing.
2. Create Visuals in Leonardo AI:
3. • Ask ChatGPT to convert your script lines into image prompts.
   • Use these prompts in Leonardo AI (Fluxscale model recommended) to generate your images.
3. Assemble and Edit in Revid AI:
4. • Upload your images in a logical sequence.
   • Add narration and background music.
   • Generate the video and review the flow, making adjustments as needed.

🔍 Observations & Tips

From our testing, here are some key takeaways:

• Script is King: The initial script used determines how good the final video turns out. Hence, manually refining the first script draft given by AI is a crucial step to keep in mind. 
• Choose your model wisely: Using Leonardo AI's Fluxscale model is a pro tip for saving credits without a major drop in visual quality.
• Sequence is Key for Revid AI: To get a logical video, ensure your uploaded images follow a clear narrative or sequence. Random image order leads to a confusing output.
• Need of human eye still prevails: While AI does the heavy lifting, the process still requires human judgment for pacing, image selection, and final edits to ensure consistency.
• Other Tools Fell Short: We tested other AI video generators, but none matched Revid AI's effectiveness in turning a series of static images into a compelling, narrated video for this specific workflow.

📽️ YouTube Video Link

Watch our full video tutorials to see this workflow in action.

📈 Content Strategy & Expansion

Once you've mastered this workflow, you’ll be all set to scale up with the following steps!

• Batch Creation: Dedicate a session to creating 5-10 scripts on different topics. Then, generate all the assets in one go to create a week's worth of content efficiently.
• Experiment with Styles: Swap out Leonardo AI for another image generator or try different voice styles in Revid AI to develop a unique aesthetic for your channel.
• Template Your Process: Create a standard structure for your scripts and a consistent visual style to build a recognizable brand with minimal effort.