Best ORMs in 2026
Libraries for querying your database from application code. 3 tools, ranked by live GitHub and npm numbers.
| Tool | Pricing | GitHub stars | npm a week | Last release |
|---|---|---|---|---|
| Drizzle ORMTypeScript ORM that maps closely to SQL, for Postgres, MySQL and SQLite. | Open source | 36k | 31M | 2 weeks ago |
| KyselyType-safe SQL query builder for TypeScript. It is not a full ORM. | Open source | None | None | None |
ORM comparisons
How to choose
An ORM lets you read and write database rows from your application code, with types that match your tables. Some stay close to SQL and add types on top. Others give you a higher-level API and tools such as a schema language, migrations and a data browser.
Choose by how you like to work. If you think in SQL, a thin ORM keeps you close to it. If you want more done for you, a fuller toolkit saves time. Check that it supports your database and the runtime you deploy to.
What to compare
- How close it stays to SQL
- Databases supported
- Migration tools
- Type safety
- Support for serverless and edge runtimes
- License, and any paid services
Questions
Do I need an ORM?
No. You can write SQL directly with a database driver. An ORM adds types, migrations and less repeated code, which helps as a project and team grow.
Is an ORM slower than writing SQL?
It adds a small amount of work for each query. For most apps the difference does not matter, and most ORMs let you write raw SQL for the queries where it does.