Privacy & Security

How MiniPx keeps your files private

Most image compression tools work by uploading your file to a server, processing it remotely, and sending back the result. During that round trip, your original photo exists on infrastructure you do not control — often for minutes, sometimes longer.

MiniPx takes a fundamentally different approach. All image processing runs in JavaScript directly inside your browser tab using the Canvas API and Web Workers. When you compress a photo, resize an image, or convert between formats, the work happens on your CPU — not on a remote server. Your files never enter a network request.

This architecture is not a premium feature or an optional mode. It is the only way MiniPx works. There is no server-side fallback, no cloud processing for larger files, no optional upload path. Every operation, from a 100KB icon to a 25MB camera RAW photo, runs locally.

What this means for you

Personal photos stay personal. Compress holiday snaps, family photos, or selfies without a copy existing on someone else's infrastructure.

Sensitive documents stay secure. Passport scans, ID photos, medical images, tax documents, contracts — process them knowing nothing leaves your device.

Business files stay confidential. Product photos before launch, internal presentations, unreleased marketing assets — compress without leaking to third parties.

EXIF metadata gets stripped. Every compression removes GPS coordinates, device info, and timestamps that are invisibly embedded in your photos. Your output files contain no hidden personal data.

GDPR and UK data protection

The architecture removes the hard part of GDPR rather than satisfying it by paperwork: because no image file is transmitted to or stored on our servers, there is no image data processing to disclose, no processor agreement to sign, and no breach surface for the content you compress. That claim is about your images, and it holds without qualification.

The website is a separate question, and it would be dishonest to fold it into the same sentence. We run Google Analytics (property G-CPKR150KT7) for page-view measurement. It sets first-party cookies (_ga, _gid) and processes your IP address and a pseudonymous identifier, which is personal data under GDPR and the UK Data Protection Act. There are no advertising cookies, no cross-site tracking pixels, and no session replay, but we are a data controller for that analytics data and we do not claim to be exempt from regulation because of it. What we can say precisely: we never see, receive, or store the images you process. Full disclosure, including lawful basis and how to exercise your access and erasure rights, is in our Privacy Policy.

Sector-specific use cases

Earlier versions of this page told regulated professionals what their compliance obligations were. That was overreach on our part. We can tell you precisely how the software behaves, and we have every incentive to get that description right. We cannot tell you what your regulator, your DPO, or your professional body will make of it, and you should be suspicious of any vendor who does. So what follows is the architectural fact, and the assessment stays where it belongs — with you.

The fact, once, for all four sectors. Image files opened in MiniPx are decoded, processed and re-encoded by JavaScript in your browser. They are not uploaded, not stored on our infrastructure, and not transmitted to any third party. There is no server-side processing path for image content and no optional upload mode. What does leave the browser is set out in our Privacy Policy: Google Analytics events carrying anonymous operation metadata, and, if you use the AI tools, a model-file download from a third-party host that sees your IP address but not your file.

NHS and healthcare. Patient images, clinical photographs, referral documents and scan results fall under NHS data governance standards, and uploading them to a server-side compression tool creates a processing relationship your organisation has to account for. MiniPx does not create that relationship for the image content, because the content never reaches us. Whether that changes your DPIA position, your DSPT return, or your Caldicott assessment is a judgement for your information governance lead — bring them this paragraph rather than our conclusion.

Education. Schools and universities handle student photos, exam scripts, safeguarding records and SEND documentation, and processing children's data attracts the higher standards of the Age Appropriate Design Code. The relevant fact is that no student image is transmitted to us. Whether MiniPx nonetheless belongs on your record of processing activities is a decision for your DPO, not for us — note that the analytics and AI-model disclosures above are the parts that would inform that decision.

Legal and professional services. Solicitors, barristers and accountants handle privileged and confidential client material, and sending a contract scan or a court filing image to an external server puts that material in a third party's hands. With MiniPx the document is never disclosed to any third party, because it never leaves the device. Whether privilege or your confidentiality obligations are affected in your specific matter is a question for your firm's risk or compliance function, and we are not in a position to answer it for you.

Government and public sector. UK departments classify material under the Government Security Classifications policy, and OFFICIAL-SENSITIVE material is not to be processed on external services without assessment. The fact we can supply for that assessment is that the material stays on the user's device throughout processing. The assessment itself, and whether your departmental policy permits the site at all, is your security team's call.

Verify it yourself

You do not need to take our word for it. Open your browser's developer tools (F12 on desktop, or use your browser's inspection mode), switch to the Network tab, then compress an image. You will see the page load resources (HTML, CSS, JavaScript) and Google Analytics events, but zero outgoing requests containing image data. Or simply disconnect from the internet after the page loads — compression, resizing and format conversion still work.

The compression logic uses the browser's native Canvas API for JPEG/PNG/WebP encoding and jsPDF for PDF creation. No proprietary server-side processing and no third-party image service ever receives your file.

There is one network request worth naming rather than glossing over, because you will see it in that Network tab and we would rather you saw it here first. The AI tools (background removal, upscaling, face detection) run their model on your device, but the model file is not ours and is not bundled with the page. The first time you use one, your browser downloads it directly from huggingface.co or storage.googleapis.com — around 44 MB to 109 MB depending on the tool, cached afterwards so it happens once. That host sees your IP address and user agent, the same as any file download. It does not see your image, because your image is not part of the request: the model travels to your file, never the other way round. If that transfer is not acceptable in your environment, every non-AI tool on MiniPx works without it.

Comparison with other tools