Catalogue
22 formats.
Each one written to its own specification, so the file you get behaves like a real one.
Archives
When you are testing transfer speed or a hard byte ceiling rather than a parser, an archive or a plain byte stream is the cleanest instrument.
Data
Data files get parsed, not just stored, so a truncated or malformed file fails for the wrong reason. These parse cleanly at any size, which keeps your test honest about the limit you meant to check.
Documents
Document uploads are where size limits bite first: a mail server, a CMS media library, a contract portal. These formats carry real structure, so a padded file still opens where a real one would.
Images
Image forms usually validate twice — once on byte size, once on decoded dimensions. Every image here carries genuine pixel data, so both checks see what they expect.
Media
Media pipelines probe a file before accepting it. These carry valid headers and, for audio, real samples — so ffprobe and friends read them rather than rejecting them outright.
System
Security scanners and upload filters treat executables, scripts and disc images differently from everything else — often by extension alone. These are built to be correctly recognized as what they claim to be, without ever shipping code that actually runs.
Questions.
Which format should I use to test upload limits?
Match whatever your users will actually upload. If you only need to hit a byte ceiling regardless of format, ZIP or BIN work against almost any accept-type.
Are these real, valid files?
Yes. Each format is written to its own specification — a PDF opens in Acrobat, a PNG decodes to real pixels. The extra bytes are padding inside a valid structure, not junk appended to the end.
Do I need to sign up or upload anything?
No. Every file is generated locally in your browser and handed to you as a direct download — nothing is uploaded, and there's no account required.