About Glowpixel Labs
Last updated:
Glowpixel Labs is a set of free image tools that run in the browser. The website was created to provide fast, reliable, and accessible utilities for everyday image tasks without forcing visitors to navigate complicated software installations, paywall subscriptions, or privacy compromises.
What Glowpixel Labs Is
Direct utility without installations, paywalls, or privacy compromises.
Everyday Utilities
Glowpixel Labs is a set of free image tools that run in the browser. The website was created to provide fast, reliable, and accessible utilities for everyday image tasks without forcing visitors to navigate complicated software installations, paywall subscriptions, or privacy compromises.
Device Processing
Every tool on the site is built to execute locally on your computer, tablet, or phone. Whether you need to resize a graphic, convert formats, compress files, or apply adjustments, the processing happens directly on your machine.
Focused Environment
Glowpixel Labs is offered in English only, providing a focused, uncluttered environment designed for direct utility. We do not bombard you with deceptive download buttons, popups, or aggressive paywalls.
The Local Processing Architecture
Why in-browser execution fundamentally outperforms conventional server uploading.
Remote Server Processing
The traditional approach to web-based image manipulation relies heavily on remote server processing. On most websites, when you select a photograph, the file is uploaded over the internet to a distant data center, processed on remote servers, and then downloaded back to your browser. This conventional model creates unnecessary delays, consumes network bandwidth, and introduces significant privacy concerns because your personal images must pass through third-party servers.
Images Never Uploaded
Glowpixel Labs operates on the opposite principle: images are processed on the device and never uploaded. Using modern web standards—such as HTML5 Canvas, OffscreenCanvas, and Web Workers—the tools perform pixel-level operations inside your browser memory. When you drag an image onto a tool, the browser reads the file locally, computes transformations through client-side scripts, and prepares the output file directly in device storage. Because your files never leave your device, you can work on sensitive images, personal photos, and confidential assets with complete confidence that no remote server ever receives or views your content.
Model and OCR Downloads
Background removal and upscaling fetch their model weights when you first start the tool and keep them in browser Cache Storage. The primary copies are served from this site; if a model file is unavailable, the tools may fetch the same public model from Hugging Face. Image to Text similarly loads its self-hosted Tesseract.js runtime and only the language data you select. These requests contain no image pixels, filenames, or extracted text. Inference and recognition run on your device.
Privacy, Accounts, and Analytics
Engineered from the ground up to respect visitor privacy completely.
Glowpixel Labs is structured to respect visitor privacy completely:
No Accounts
You never need to register, provide an email address, create a password, or manage a user profile to access any tool on the site. Every feature is available immediately to every visitor.
No Analytics
We do not install analytics packages, tracking pixels, or behavioral telemetry tools. We do not track page views, feature usage, or navigation pathways across the website.
No Site Cookies
The site sets no cookies itself. Instead of cookies, the site uses standard browser localStorage for two functional choices: Interface theme stored under the key gp-theme to remember whether you selected dark mode or light mode across different pages, and advertising consent stored under the key gp-consent. If persistent storage fails after a decline, gp-consent-deny in sessionStorage preserves that choice for the current tab. These keys remain on your device and can be cleared through your browser settings.
Infrastructure and Operations
How the project is hosted, maintained, and operated transparently.
Edge Infrastructure
Glowpixel Labs is hosted on GitHub Pages behind Cloudflare for fast, secure static delivery worldwide. When you load pages, style sheets, and scripts, your web browser connects directly to Cloudflare and GitHub Pages edge servers. In accordance with standard internet networking, Cloudflare and GitHub Pages see IP addresses and request metadata in server logs to route web requests, mitigate network threats, and maintain service stability. Glowpixel Labs does not compile or analyze these logs to track individual users.
Consent-Gated Advertising
The website is free to use. A deployment may enable Google AdSense to help cover operating costs, but the Google script is requested only after a visitor affirmatively accepts advertising. If advertising is declined or no choice is made, that script is not loaded. When enabled and accepted, Google and its partners may use cookies and similar technology under their own policies. Images selected in a tool are never supplied to advertising code.
How to Get in Touch
If you have questions, feedback, or suggestions regarding the tools or how the site works, you can contact us by email at hello@glowpixellabs.com. This is a direct mailto link; the site does not operate a web contact form. Because all communications and tools are in English only, please write your message in English. As a reminder, because images are processed on your device and never uploaded, please never attach or send images by email.
Open-Source Software and Models
Third-party components delivered by the production site.
- fflate — Version 0.8.3; MIT; creates ZIP downloads and compresses image data in generated PDFs.
- @jsquash/avif 2.1.1, @jsquash/jpeg 1.6.0, @jsquash/oxipng 2.3.0, @jsquash/png 3.1.1, and @jsquash/webp 1.5.0 — Apache-2.0; decode and encode images in the converter and compressor.
- heic-to — Version 1.6.5 with libheif 1.23.5; LGPL-3.0; lazily decodes HEIC and HEIF images in the browser.
- onnxruntime-web — Version 1.30.0; MIT; runs the background-removal and upscaling models on the device in background workers, with its WebAssembly and WebGPU runtime files served from this site.
- MODNet
model_quantized.onnx— shipped ONNX model; Apache-2.0; removes portrait backgrounds. - Real-ESRGAN general x4v3 (SRVGGNetCompact)
realesr-general-x4v3.onnx— shipped ONNX model; BSD-3-Clause; upscales images by 2× or 4×. - IS-Net general use
isnet-general-use-q8.onnx— downloaded on demand from the recorded SacredNoir model repository rather than bundled inpublic/models; Apache-2.0; powers the Background Remover’s Pets and objects mode. - @mediapipe/tasks-vision — Version 0.10.21, with the BlazeFace short-range model
blaze_face_short_range.tflite(float16/1); Apache-2.0; finds faces locally for Blur & Pixelate. - tesseract.js — Version 7.0.0, with its OCR core and tessdata_fast language files; Apache-2.0; extracts text locally in Image to Text.
- @fontsource-variable/instrument-sans, @fontsource-variable/hanken-grotesk, and @fontsource/geist-mono — Version 5.3.0; SIL Open Font License 1.1; provide the self-hosted interface fonts.