Privacy Policy

This policy applies to the Pixelify website and its tools (the AI Image Upscaler and the Background Remover). It is written to be verifiable: where we describe data handling, we describe the behaviour of the actual application, and you are welcome to check it against the source.

Pixelify has no application backend or user accounts. A small external counter records aggregate site visits; the limits of that counter are described below. The following data is not collected by Pixelify or sent to the visit counter:

• Your images. Files you drop, paste or select are read into memory on your device. No pixel, thumbnail or file name is transmitted to us or to any image-processing server.
• Account data. There is no registration, login, email capture or password. We could not hold account data even if we wanted to.
• Payment data.The tools are free. There are no purchases and no payment processors.
• Behavioral analytics or profiles.We do not record clicks, tools used, images processed, session replays, or build profiles across pages or sites. The site counter only records an aggregate visit total and uses a short-lived deduplication hash (see "Network requests").
• Selling data. We have no bulk data to sell and no advertising relationship that involves your content. The ad partner only sees what ad networks see on any site (see "Advertising").
• Cookies.Pixelify and the visitor counter set no cookies. Embedded ad providers may set their own cookies; see "Advertising".
• Location and device fingerprints.No geolocation requests, no fingerprinting scripts.

A few things are stored locally, by your browser, on your device. We never receive them; they exist only as your browser's site data. You can delete all of it at any time via your browser's "clear site data" control for this site, which resets Pixelify to a first-visit state.

• Preferences (local storage). Your theme choice (dark / light / system) and the sound-on/off setting for the completion chime. Two small key-value entries, no identifiers attached.
• Model cache (Cache Storage). Neural model files you have used (roughly 2MB to 67MB for the upscaler, about 40MB for the background remover) are cached so they do not re-download on every visit. These are stock model weights, not data about you.
• Images, session only. Your images live in memory for as long as the tab is open. Closing or refreshing the tab discards them. Nothing is written to persistent storage; a legacy local history feature existed in earlier versions and has been removed and actively deleted on load.

Honesty requires listing these, because "runs locally" describes your images, not the ordinary requests a browser makes to load any website. Pixelify makes the following requests. None carry your images; some providers receive standard network metadata such as an IP address and user agent when serving a request:

• The site itself. The application files are served by our hosting provider, Vercel. Like every web server, Vercel receives standard request logs (IP address, user agent, time, page requested) for delivering the site. See the Vercel Privacy Policy.
• Model files. Upscaler models are fetched from the same domain as the site and then cached by your browser. The background remover's Chopper model is downloaded through Transformers.js from the Hugging Face model hub. It is a one-time ~40 MB file that is then cached. These requests contain a URL and nothing about your images.
• Open-source libraries. ONNX Runtime Web (the inference engine) and its WebAssembly binaries are loaded from jsDelivr, a public content-delivery network. jsDelivr serves the library file; it never sees your images.
• Fonts. Typefaces are served by Google Fonts. Google receives the standard request data involved in serving a font (IP address, user agent). If you prefer to block it entirely, the site remains fully usable with system fonts.
• Site visit counter. When a page loads or the route changes on pixelify.me, the browser sends a shared site-level counter key and standard request metadata to the Page Views API(its source is public). Its documentation says it derives a short-lived SHA-256 hash from the IP address and user agent to group repeat visits for about 30 minutes, and does not store the raw IP address or set cookies. We send the same counter key for every route, so route names are not included in the counter data. The resulting aggregate count is shown in the floating visitor badge on all pages.

That is the complete list of requests the site itself makes. Advertising (AdSense) adds requests described in its own section below, including third-party cookies.

Pixelify is free and has no accounts; the ads keep it that way. The site uses Google AdSense for advertising:

• What loads. AdSense scripts (from googlesyndication.com and related Google domains) are embedded in the page. They run Google's code, not ours.
• Cookies. Google uses cookies (including DoubleClick/DART cookies) to serve ads based on your prior visits to this and other sites. These are set and controlled by Google. Pixelify never sees the values. You can opt out of personalized advertising in Google's Ads Settings, or via aboutads.info.
• Images stay local. The ad scripts have no relationship with the AI tools. Your images are still never uploaded or shared with any ad network; there is no cross-talk between files in your workspace and the ad slot.
• Where ads appear. Ads are limited to content pages (articles, blog, landing pages), not inside the AI workspaces.

For details on how Google handles ad data, read Google's Advertising policies.

Summarised: Vercel (hosting), Page Views API(aggregate site visit count), Hugging Face(background-removal model weights), jsDelivr (inference libraries), Google Fonts (typefaces) and Google AdSense(advertising). Each processes the limited request data described above. We have no commercial relationship with them that involves your content. The counter receives a shared site-level key and ordinary request metadata; no image or editing activity is sent to it.

External links inside our articles lead to third-party sites. Their practices are governed by their own policies, not ours.

Pixelify does not keep user profiles or raw visitor identifiers. The counter provider maintains aggregate totals and says its temporary deduplication hash expires after about 30 minutes. For questions about that processing, you can also contact us using the contact details below. The data stored by your browser remains under your control:

• Clear site data for this origin in your browser settings to erase preferences and model caches instantly.
• Block notifications: the optional desktop "done" alert uses the browser notification permission, which you can grant, deny or revoke at any time in browser settings. We ask only when you start a job.
• Use private/incognito browsing and nothing persists at all after the tab closes.

If you have a statutory data request about the counter, contact us and we will help direct it to the service that processes the count. Pixelify does not have an account or dashboard for that service and cannot edit its aggregate records directly.

Pixelify is not directed at children under 13 (or the equivalent age in your jurisdiction). The site has no account system; the aggregate site visit counter described above does not create a user profile.

If this policy changes, the updated version will be published on this page with a revised date at the top. We will call out changes that alter what leaves your device.

Questions about privacy, or requests for clarification about anything above, are welcome. The fastest route is a GitHub issue: