In most SQL databases, once you have a connection handle, you can allocate an arbitrary number of statement handles, and each SQL "statement" can be executed independently, has an associated query, parameters, a result set, and so on.
It is entirely ordinary with an API like that to prepare a statement, bind parameters and columns, and execute and fetch the results. You can then reuse a statement in its prepared state, but usually with different parameter values, as many times as you want within the same session.
The performance advantage of doing this for non-trivial queries is so substantial that many databases have a server side parse cache that is checked for a match even when a client has made no attempt to reuse a statement as such. That is easier if you bind parameters, but it is possible for a database to internally treat all embedded literals as parameters for caching purposes.
It is entirely ordinary with an API like that to prepare a statement, bind parameters and columns, and execute and fetch the results. You can then reuse a statement in its prepared state, but usually with different parameter values, as many times as you want within the same session.
The performance advantage of doing this for non-trivial queries is so substantial that many databases have a server side parse cache that is checked for a match even when a client has made no attempt to reuse a statement as such. That is easier if you bind parameters, but it is possible for a database to internally treat all embedded literals as parameters for caching purposes.