Creating Talking Videos with AI: A Step-by-Step Guide

AI to the Rescue: Crafting a Quick Product Video

Hello, this is Jey from DaaleelaB by NaaveelaB. I’m a designer and AI artist based in South Korea, selling my creative products on Amazon’s U.S. marketplace. Recently, I found myself in urgent need of a product video but couldn’t find a videographer who could produce something I loved within a reasonable budget. So, I turned to my trusted AI tools and created a temporary product video on my own.

Since this is only the third issue of my newsletter, I’m still figuring out what topics resonate most with readers like you. If you’re interested in creating your own products or building a brand—or even if you simply want to make social media videos—you might be curious about how AI tools can help. For this issue, I’ll explain how I used AI tools to create a talking product video for my Amazon listing.

While learning AI tools like MidJourney from scratch is an option, I find it more practical to focus on completing small, specific tasks and picking up the relevant tools and features along the way. Here, I’ll break down how I approached the task of creating a “talking video” with AI, step by step:

1. Search for a natural-looking portrait on Pexels with a suitable background.

2. Use MidJourney’s "Describe" function with this image to generate a prompt.

3. Input the prompt and upload the Pexels image as a reference in MidJourney to generate a similar-looking portrait.

4. Take the resulting image, add a script, and upload both to Hedra.com to create a talking video.

Hedra is extremely easy to use—just write your script, upload your image, and generate the video (see the screenshot for details). The tool even works for animals, so if you ever need a talking cat in your video, Hedra’s got you covered! For my Amazon product video, I used Hedra to create a talking video for the intro and outro, while extracting only the audio from the middle section to overlay on product photos (I used Adobe Premiere to separate the audio).

Refining the Process: Combining Tools for a Polished Result

To create the final video, I zoomed in on product photos using Kinemaster app, a feature available in most video editing tools. Only for the first zoom effect, I used a generated video by Flair.ai from a product shot. With just five static images (four product shots and one AI-generated person), I was able to create a simple but effective temporary video. Of course, I’m planning to produce a more polished product video soon, ideally with professional filming, as I believe that’s what potential buyers want to see.

Hedra also allows you to upload pre-existing photos—or even take one on the spot—to create talking videos without involving MidJourney. If you have photos of your products with people already in them, this could save you an extra step. However, always ensure you have the copyright for the photos you use to avoid any issues.

I hope my method inspires you to try creating videos yourself—or at least shows you how simple and accessible the process can be. If you’d like me to go into more detail about any specific step, feel free to leave a comment or send me a message. For now, I’ll brainstorm ideas for the next newsletter and look forward to sharing something new with you soon!