How expensive is the NOT IN/NOT EXISTS SQL operation?

Question

PostgreSQL

I'm wondering how expensive SQL queries with NOT IN or NOT EXISTS will be?
Let's say there are 2 tables
items and using_items
In the items table there are 12 million records, in using_items 11 million. Of course, the number is hypothetical and unlikely.

SELECT t.id, t.name 
FROM items t
WHERE (t.type = 'some_item_type') AND t.id
      NOT IN (
          SELECT w.id_item
          FROM using_items w
         )
ORDER BY t.name ASC;

or

SELECT t.id, t.name 
FROM items t
WHERE (t.type = 'some_item_type') AND 
      NOT EXISTS (
          SELECT w.id_item
          FROM using_items w
          WHERE t.id = w.id_item
         )
ORDER BY t.name ASC;

How bad will everything be in each of the requests? If it is very bad, what are the possible solutions?
Or will you have to split the using_items table into tables by the type attribute (using_items_type1, using_items_type2)?
Thank you.

Answer 1

SQL JOIN

Second from the left.