How we decide a tool is ready
A format name is added to the public directory only after the conversion exists. This page explains the checks and the limits of that evidence.
Functional checks
We use representative valid files, damaged or mislabeled inputs, empty files, boundary sizes, multiple selections where supported, and output reopening. Image checks include dimensions, transparency, and high-resolution memory limits. PDF checks include multi-page files, page ranges, and protected or malformed inputs. Archive checks include duplicate names and expansion limits.
Browser checks
Browser media codecs differ by operating system. A converter that uses a browser decoder can therefore work in one current browser and reject the same codec in another. Tool pages disclose this where it applies. Simulated responsive testing does not replace physical-device or assistive-technology testing.
Privacy checks
Conversion modules read selected files through browser APIs. The converter does not POST file contents to our server. Analytics must never contain filenames, document content, extracted text, or converted output. Browser extensions, downloaded files, and the visitor’s own device remain outside the site’s control.
Publishing rule
We do not publish placeholder routes for conversions that are merely planned. Known quality loss, metadata removal, animation loss, page limits, and browser dependencies are described on the relevant tool page.
Report a problem
If a published tool fails, send the tool name, browser, operating system, approximate file size, and error message through the contact page. Do not send a private file.