JDBC vs Hibernate vs JPA vs Spring Data JPA

Four Names, Four Layers

Beginners often mix these up. They are layers stacked on top of each other:

Your code (ProductService)
        |
        v
 Spring Data JPA   -> interface ProductRepository extends JpaRepository<Product, Long>
        |             (generates repository code for you)
        v
 JPA (Jakarta Persistence 3.1) -> the STANDARD API: @Entity, EntityManager, JPQL
        |
        v
 Hibernate ORM 6   -> the IMPLEMENTATION of JPA (generates SQL, caching, dirty checking)
        |
        v
 JDBC              -> low-level Java API that sends SQL to the database
        |
        v
 MySQL / PostgreSQL / H2 driver -> database

Comparison

	What it is	You write	Package
JDBC	Low-level Java API for SQL	SQL strings, ResultSet loops	`java.sql`
JPA	A specification (rules + interfaces), no code that runs by itself	Annotations, EntityManager calls, JPQL	`jakarta.persistence`
Hibernate	The most popular JPA implementation, plus extra features	Same as JPA (optionally Hibernate-only APIs)	`org.hibernate`
Spring Data JPA	Spring library on top of JPA that removes repository boilerplate	Repository interfaces only	`org.springframework.data.jpa`

Note: JPA used to be called Java Persistence API (javax.persistence). Since Jakarta EE 9 it is Jakarta Persistence and the package is jakarta.persistence. Spring Boot 3 and Hibernate 6 use only the new jakarta.* package — old tutorials with javax.persistence will not compile.

Same Task at Each Layer

Loading one product by id:

// JDBC: you write everything
try (PreparedStatement ps = conn.prepareStatement("select id, name, price from product where id = ?")) {
    ps.setLong(1, 42L);
    try (ResultSet rs = ps.executeQuery()) {
        if (rs.next()) {
            product = new Product(rs.getLong("id"), rs.getString("name"), rs.getBigDecimal("price"));
        }
    }
}

// JPA (Hibernate underneath): one call, SQL generated
Product product = entityManager.find(Product.class, 42L);

// Spring Data JPA: you only declare an interface
Optional<Product> product = productRepository.findById(42L);

Other JPA Implementations

Because JPA is a standard, other implementations exist: EclipseLink (the reference implementation) and OpenJPA. Spring Boot uses Hibernate by default, and almost every job posting that says "JPA" means Hibernate.

Tip: Write code against the jakarta.persistence API wherever possible. Use Hibernate-specific APIs (org.hibernate.*) only when you need a feature JPA does not have.