Privacy Policy

ezmod.ai does not send your images to a server. Every image operation finishes inside your browser, and no address that would accept an image exists in the code. Even so, here is the complete account of what this service does send and receive.

Images

We do not collect them. Nothing — the image you open, the capture data (EXIF) inside it, or the file the tool produces — is sent to a server or stored on one. Your browser creates the result file and your browser hands it to you.

Accounts

There is no sign-in anywhere in this interface. A Google sign-in endpoint from an earlier version is still present in the deployment, so where that configuration remains in place, entering that address directly lets you sign in with a Google account. Only in that case is the following stored in an external database (Neon Postgres): the name, email address and profile image URL that Google provides, an identifier for your Google account, and your account's role, suspension state, creation time and last sign-in time. We do not store Google's access token or refresh token, and the sign-in session itself is kept in an encrypted cookie.

That endpoint is scheduled for removal, and this section goes away once it is gone.

Usage analytics

We measure page views and page performance with Vercel Analytics and Vercel Speed Insights, and we also record what happens in the tools. About an upload: which tool was used, how the images arrived (dropped, pasted from the clipboard, chosen from a file dialog, or carried over from another tool), how many at once, how many of them were turned away, and whether it was the format, the size limit or the limit on how many fit in one batch that turned them away. About a job: whether it finished or failed, how many images got through and how many did not, roughly how long it took as one of a few coarse ranges, and — where it failed — which category the failure fell into and which format the processed image was (PNG, JPEG, WebP, AVIF, GIF, or anything else as a single catch-all). And about what comes after: whether results were downloaded, how many, and whether as one file or a zip; and whether the upscaling model had to be fetched or was already held from earlier, whether fetching it succeeded, and roughly how long that took. None of that carries filenames, image contents or capture data (EXIF), and none of those ever leave your browser. Neither Vercel product uses cookies, and neither stores IP addresses. What is stored besides: the address visited and the referrer, the visitor's region (city-level for Analytics, country-level for Speed Insights), the browser and device type, and performance measurements. Analytics tells visits apart using a hash derived from the request, which is discarded after 24 hours. Neither tracks you across sites.

What is stored in your browser

So the service starts quickly and works without a connection, the files the app is made of (code, styles, fonts) are stored in your browser's cache. On top of that, your interface settings (theme, sidebar width, compression options, gallery view mode) are stored in your browser so your next visit starts the way you left it. If you use upscaling, the model files you downloaded (about 67 MB each) are stored too, so they do not have to be downloaded again. Images are not stored. Clearing this site's data in your browser removes all of it.

Requests that leave your browser

When you use upscaling, your browser downloads the AI model file from Hugging Face and the files needed to run it from jsDelivr. All that leaves is a request for those files — no image travels with it. Both services may log the connection under their own policies.

Contact

Questions about this policy go to needmorecache@gmail.com.

Changes

If this policy changes, we update this page.

Last updated: September 20, 2026