snapCopy Enters Testing: One Photo, Better Captions for Sharing

Today, Station Cat’s first small app, snapCopy, officially entered testing.

The idea is simple.

Sometimes you are having breakfast and take a photo of coffee and bread.
Sometimes you are out for a walk and notice a street that looks just right.
Sometimes the cat happens to sit by the window, and the light happens to fall in the perfect place.

The photo is already there, but the last step gets stuck:

What caption should go with this image?

That is the small problem snapCopy is trying to solve.

What It Can Do Right Now

This test version is focused on a few basic features.

You can choose an everyday photo.
The app tries to recognize the scene inside it.
For example: breakfast, coffee, walking, pets, travel, outfits, street views, table scenes, and other ordinary daily moments.

Then it generates several captions that should be easier to share.

You can copy these captions or use them on X, Instagram, Moments, or other social platforms.

I want it to feel more like a small daily assistant:

Not something that writes a formal article for you.
Not something that tries too hard to manufacture a viral tone.

It should feel more like a gentle nudge after you take a photo:

“You could say it like this.”

The Most Important Things to Test

For this round, I care less about how complex the features are and more about whether the app can do the basic things well.

First, can it understand the photo?

If the photo is breakfast, it should try to understand that it is breakfast.
If the photo is coffee, it should try to understand that it is coffee.
If the photo is a street view from a walk, it should at least understand that this is not just a generic background image.

Image understanding is the foundation of the whole app.

If it misses the content of the image, the captions, recommendations, ratings, and style learning that come later will drift in the wrong direction.

Second, can the captions fit the photo?

I do not want it to generate empty sentences.

For example, when looking at a coffee photo, it could write:

“Let’s do our best today too.”

That is usable, of course, but it misses the atmosphere inside the image.

I want it to generate expressions that feel closer to the scene:

A cup of coffee, a little sunlight, and the morning slowly comes back.
Today’s breakfast task: take care of yourself first.
Found a small piece of city softness on the walk.

The caption does not need to be exaggerated. It just needs to stand next to the photo.

Third, does rating feedback help?

The test version will include rating and feedback logic.

If you like a certain caption, you can give it a higher score.
If something feels too plain, too long, or too template-like, you can give it a lower score.

Over time, the app should gradually remember your preferences.

Maybe you like short sentences, a relaxed tone, and a little humor.
Or maybe you prefer quieter, more Japanese-inspired wording that feels better for daily records.

This direction is still early, but it is an important part of where snapCopy should go next.

Why Start With On-Device AI

The current version prioritizes Apple’s on-device AI capabilities for caption generation.

There are a few reasons for that:

It is faster.
It creates less privacy pressure.
The basic usage cost is lower.
It also fits a lightweight daily tool better.

I will also leave room for cloud enhancement later.

For example, if local recognition is not accurate enough, users could choose stronger cloud image understanding.
Or if they want captions that feel more platform-aware and closer to social content, enhanced generation could help.

But this will not be the default feature at the beginning.
Cloud enhancement involves cost, privacy notices, and usage limits, so it will be tested more carefully in later versions.

What I Hope It Becomes

I do not want snapCopy to become a cold “caption machine.”

It should feel more like a small tool that understands your style.

When you photograph breakfast, it knows you like lighter sentences.
When you photograph a cat, it knows you do not want overly cute captions.
When you photograph a view from a walk, it knows you prefer something with a little looseness and calm.

It does not take your expression away. It just helps organize the thing you wanted to say but did not feel like putting together.

What Comes Next

Over the next while, I will focus on adjusting:

Photo scene recognition accuracy
Manual scene selection when confidence is low
Caption generation quality
Copy and share flows
Rating feedback
Personal preference learning
TestFlight user feedback
The website download page and waitlist

If testing goes well, later versions will gradually add:

More scenes
More languages
Caption styles better suited for X, Instagram, and Japanese social platforms
Cloud-enhanced generation
The Android waitlist
Preparation for official release

If You Want to Test

snapCopy is currently in TestFlight testing.

If you often take photos of breakfast, coffee, cats, walks, travel, street views, or strange little moments you still want to share, you can join the waitlist.

When the next round of testing opens, I will share the update through the website.

This is only Station Cat’s first small app.
It is still early, and it will definitely do many not-so-smart things.

But if it can help someone spend one less minute stuck on a caption and share one more photo they wanted to post, this small tool has already started to mean something.

The cat will keep looking out the window.
The table will keep appearing.
Strange ideas will keep popping up.

snapCopy will keep growing too.