Automate Image Labeling / File Naming with OpenAI GPT

If you take a lot of screenshots, you might be familiar with this issue: you follow a step-by-step tutorial, capture screenshots, and end up having a bunch of files named like Screenshot1, Screenshot2, and so on.

These names don’t really tell you what each image shows.

Sure, you can use a preview to see the content, but wouldn’t it be better if the file name actually revealed what’s inside?

Why Meaningful File Names Are Helpful

• You can sort, pick, and choose images for your tutorial more easily.

• You can use AI to understand the file’s content, even without visually checking it.

• Meaningful file names are much better for SEO.

• It’s easier to search for images when they have descriptive names.

One Way to Do This on a Mac

The goal is to simply right-click any image or a group of images and let AI generate (and replace) the filenames based on what’s in the images.

We’ll use:

• Apple’s Mac Shortcuts app

• OpenAI gpt-4o-mini with vision capabilities

In the end, it looks like this:

When I right-click a screenshot named “Screenshot-4” and choose the filename generator shortcut:

The file gets renamed according to its content (in this example, “abstract-hand-logo”):

The AI File Naming Shortcut

To make this work, we create a new shortcut that sends the file to OpenAI along with a prompt like:

“Suggest a filename for this image. Answer with the filename only, no comments!”

Here’s an overview of the Shortcut workflow:

The core of the shortcut is the call to OpenAI’s vision feature.

Note that we need to pass two pieces of information: the prompt and the image (as a base64-encoded file).

More about OpenAI for vision can be found here: https://platform.openai.com/docs/guides/vision

You can also use AppleScript or JavaScript automation to do this directly, without clicking through the whole shortcut in the Shortcuts app (though that’s not covered in this tutorial).

What’s Next

You can use a similar approach not just to rename files, but also to label important images for tutorials, blog posts, or any place where you often need to sort through images and Apple’s search isn’t enough.

You could also use the same technique to add tags or comments to files, making them easier to find later.