Your question is really broad. I suggest reading e.g. the Wikipedia article on open access to get some sense on what it's about.
Traditional journals operate by subscriptions. They publish articles, and would-be readers must subscribe to be able to read them (or they can buy individual articles, which is similar). This model has been criticized for making it difficult to disseminate academic knowledge. After all, if you can't pay, you can't read the paper either and so won't be able to build on that work. Since academic publishing costs money, someone has to pay, and OA proposes to make that "someone" the authors. The authors pay the publishers, who then makes the articles available for free.
The main problem with OA is that it introduces a conflict of interest. Since the publisher makes money every time it accepts a paper, it (and by extension the editorial board) is incentivized to accept papers. This has led to so-called predatory open access publishers, who basically accept everything that's submitted with no thought for quality control. Subscription journals cannot do this because if they publish junk people won't subscribe, but OA journals can.
That is not to say that all open access articles are junk or that all open access publishers are predatory, but you can imagine the tension. Not helpful is that some people consider X predatory while others do not. Hindawi and MDPI are examples. Both were listed by Jeffrey Beall (the authority on predatory OA publishing before he retired) as predatory at some point, but they also had vocal defenders.
Practically speaking, open access papers do have more usage than subscription ones, but whether that translates to more citations is hotly debated. Whether or not to publish open access is up to you, and possibly your funding agency.