how does this helps a multi language site ?
It does nothing to help a multi-language website per se. What it does is that it notifies the web browser that the web page is encoded in UTF-8. See this article for more information.
At a trivial level, If it's set across the front end, backend (I'm presuming there's some form of Content Management System that's used to enter text, etc.) and database collation it will allow non-ASCII characters to be entered, stored and subsequently output correctly.
That said, this is really only skimming the surface.
UTF-8
is an Unicode character encoding (charset) which covers all of the characters known at the human world as outlined here. For example the ISO 8859-x
doesn't cover them, it only covers the Latin characters (a-z and so on). If you want your webapp to be able to display and handle non-Latin characters, then you really need an Unicode charset, otherwise youw webapp will go Mojibake.
The article of Joel is excellent to start with and if you have a Java background you may find this article useful as well, it does not only explain the use of charsets in detail, but it also contains practical examples and complete solutions how to get it to work properly in a Java (JSP/Servlet) webapplication: Unicode - How to get the characters right?