About ConvertUnlimited

Short answer

What the product is

ConvertUnlimited groups its tools into three areas: image tools (format conversion between JPG, PNG, WebP, AVIF, and HEIC, plus compression, resizing, background removal, watermarking, and metadata removal), PDF tools (merge, split, compress, and conversion to and from images), and developer and data tools (JSON formatting, CSV conversion, Base64 encoding, hash generation, JWT decoding, regex testing, diff checking, QR code generation, and UUID or timestamp utilities).

Each tool follows the same pattern: open the tool page, choose a file or paste input, run the supported processing step, and download or copy the result. There is no account system and no required sign-up step for using a tool.

The browser-native processing model

For supported workflows, file bytes are read with the browser's File and Blob APIs, decoded and re-encoded using Canvas or a client-side library, and written back out as a Blob or object URL for download. ConvertUnlimited does not provide a server-side upload endpoint for these processing flows, so there is no backend ingest path for the file itself.

This model has real engineering tradeoffs. It removes the need to operate and scale a file-processing backend, but it inherits the limits of the visitor's own browser and device: available memory, CPU speed, and format support all vary by browser and hardware. A large PDF or a high-resolution photo batch that a server could handle easily can fail or run slowly in a browser tab, and some mobile browsers may not support a given encoder at all.

Libraries that do the processing

Several workflows depend on third-party JavaScript libraries rather than native browser APIs alone. PDF handling uses pdf-lib 1.17.1 for merge, split, and compress operations, jsPDF 2.5.1 to build PDFs from images, and PDF.js 3.4.120 (with its worker build) to render PDF pages to images. Batch downloads that bundle multiple output files use JSZip 3.10.1. HEIC photo conversion uses heic2any 0.0.4, which wraps a WebAssembly build of libheif and libde265 because browsers do not natively decode HEIC.

On the public site these libraries load from third-party content-delivery networks (cdnjs, unpkg, or jsDelivr, depending on the library), which is a network request in its own right, separate from any file you process. The background-remover tool ships a default local flood-fill mode with no added library, plus an optional 'AI Deep Cutout' mode that loads @imgly/background-removal from jsDelivr and downloads an approximately 40MB model file on first use, still running inference in the browser rather than sending the image to a server.

The privacy build

Limitations

Browser-native processing depends on browser format support, device memory, and processing speed, and outputs should be checked before relying on them, particularly downstream of a compression or conversion step. Metadata removal re-encodes supported image formats to strip embedded fields such as GPS coordinates or camera data; it does not alter or remove visible pixel content such as faces, addresses, or text already visible in the image.

The default background-remover mode works by detecting a fairly uniform background color and removing connected regions of similar color; it cannot reliably separate a foreground subject from a background of a similar or identical color when there is no visible edge between them. The optional AI mode is intended for that harder case.