"arXiv (pronounced "archive"—the X represents the Greek letter chi [χ])[1] is an open-access repository of electronic preprints and postprints[...] "

https://openaccess.mpg.de/Berlin-Declaration

Publishers do misuse the concept though. They try to stay clear of using the term when they do. They use terms like Free Access or some of the more dubious colour variants of OA that doesn't provide all the freedoms that the Berlin Declaration of OA defines.

Interestingly articles uploaded to arXiv with the arXiv.org perpetual non-exclusive license are not OA as the reader is not allowed to redistribute the paper.

This is not true and comments like this are damaging to science.

Open Access papers are still peer-reviewed and by far not all of them make it into the journal. You can't pay your way into those journals.

Of course there are shady pseudo-journals which just cash in on the fee, do not carry out peer review and just dump the paper on the internet. But any scientist should be able to tell such scam journals from serious ones.

True, some journals, like many of the Frontiers series or PLOS One, make it very hard to be rejected in peer review. As long as your paper is reasonably well written and doesn't contain falsehoods it will almost certainly make it to publication. Still, you don't "pay your way into those journals".

Granted, many of papers in those journals report mere incremental progress. But these journals are still attractive for scientists to publish in, for obvious reasons. Publishers like PLOS use those journals as cash cows to fund their higher-tier offerings.

It is fair that the author pays for publication in those journals, since the most benefit is often for the author, not the reader. For the progress of science these offerings are not so useful, unfortunately.