Search the App Store for a photo cleaner and you get dozens of results, nearly all of them free to install and paywalled the moment you try to delete anything. Here is how to tell them apart.
Free to install is not the same as free
Search the App Store for a photo cleaner and nearly every result is free to download. Install a few and the pattern becomes obvious: the app scans your library, shows you an impressive number of gigabytes it could recover, and then asks for a subscription before it will let you act on any of it.
This is a deliberate design. The scan is the hook, the number is the pitch, and the paywall arrives at the exact moment you feel motivated. It is effective and it is why the category has a poor reputation.
The three things that actually matter
Most feature comparisons in this space are noise. Three questions separate a good cleaner from a bad one.
- Where does processing happen? On device means your photos never leave. Anything that uploads is a meaningfully different privacy proposition.
- Is deletion batched and reversible? A swipe or a tap should queue, not delete. You should get a review step and a confirmation.
- Is the core loop paywalled? Check whether you can actually delete anything without paying, not just whether the download is free.
The privacy question is the one most people skip, and it is the most consequential. A photo cleaner requests access to your complete library. That is years of images of your family, your home, your documents and your private life. It is worth thirty seconds to check what the app does with that access.
Warning signs
A few patterns reliably indicate an app not worth your time.
Deletion with no review step. If a tap removes a photo immediately with only Recently Deleted as a backstop, the app is trading your safety for a snappier demo.
Vagueness about processing location. Privacy policies that discuss uploads in general terms without stating plainly whether photos are transmitted are usually vague on purpose.
Aggressive weekly pricing. Weekly is the dominant model in this category and it is dominant because it converts, not because it reflects value. A task you perform three times a year does not warrant a weekly charge.
Inflated recovery claims. Numbers that count Recently Deleted, or that assume you will delete every duplicate it found, are marketing rather than measurement.
What to look for instead
A good cleaner is boring. It shows you your photos, makes deciding easy, holds the decisions somewhere safe, and asks once before doing anything irreversible. It tells you honestly how much space you will recover, and it does not treat your library as an opportunity to upsell.
It should also respect that you might stop halfway. Progress ought to be saved, and the app should not require you to start again if you close it.
Where Stroll fits
Stroll is free, and free in the sense that matters: there is no paywall on the core loop and no subscription to clear your entire library.
It processes everything on device using Apple's photo APIs, so no photo is uploaded and no server holds one. Swipes queue in an Archive that you review before a single confirmed batch delete, which reports the exact space you are about to recover.
It is deliberately narrow. It does not do AI-powered duplicate detection or photo editing. It does the one job well and does not ask for a card.
How the free-to-download model works
The pattern is consistent enough to describe precisely. You install the app, grant full library access, and it runs a scan with a progress animation. It then reports a large number of gigabytes it claims it can recover.
That number is the product. It is frequently inflated: counting Recently Deleted, assuming you will delete every photo it flags as similar, or including videos you obviously want to keep. It arrives at the moment you are most motivated to act, immediately followed by a subscription prompt.
None of this is illegal and some of these apps work fine once paid. But it is worth recognising the mechanism, because it explains why the App Store is full of cleaners with excellent ratings and furious reviews.
Duplicate detection is less useful than it sounds
Automatic duplicate and similar-photo detection is the headline feature of most paid cleaners. In practice it is the feature people abandon first.
True byte-identical duplicates are rare unless you have imported the same library twice. What these tools actually find is similar photos, which means burst frames and near-identical shots, and deciding between those requires exactly the human judgement the automation was supposed to replace. You end up reviewing them one at a time anyway, just inside a worse interface.
Reviewing sequentially handles the same problem without pretending to be clever, because burst frames appear next to each other in a chronological library.
A short checklist
Before granting any cleaner access to your library, confirm four things.
- The privacy policy states plainly that photos are processed on device.
- Deletion is queued and confirmed, not immediate.
- The core loop works without payment.
- Recovery claims are not counting things you already deleted.
If an app fails any of those, there is another one that does not.