Hive is a Hadoop-based storage system. Hive uses a special SQL dialect (HiveQL) to operate with data and metadata. Generally, it is quite similar to SQL.
There are multiple implementations of storage systems which utilize Hive on the server-side - including Apache Spark, Impala, etc. Most of them support the standard Hive JDBC driver which is used in DBeaver to communicate with the server.
DBeaver uses a so-called Hive JDBC Uber Jar driver (https://github.com/timveil/hive-jdbc-uber-jar) which includes all necessary dependencies. You do not need to download anything - DBeaver will download everything automatically (if you have internet access).
Hive does not support referential integrity so you will not see primary keys or foreign keys. ER diagrams also do no make much sense.