Database

Where your data actually lives.

This is where your application's data actually lives, so the tradeoffs here, managed versus self-hosted, SQL versus NoSQL, serverless versus always-on, matter more than most other picks. Managed providers save operational overhead at the cost of some control and predictable pricing. Choose based on your data shape and query patterns, not just what's fastest to set up today.

MongoDB Atlas

Managed MongoDB with a genuinely generous free tier. Still the easiest on-ramp if your data is document-shaped rather than relational.

↗

Neon

Serverless Postgres with git-like branching, so every PR or preview deploy can get its own isolated copy of the database.

↗

PlanetScale

Managed MySQL (and now Postgres) built on Vitess. Good for schema branching and non-blocking migrations at scale.

↗

Supabase

Hosted Postgres with auth, storage, and realtime bundled in, plus auto-generated REST/GraphQL APIs. The fastest way to get a real relational DB behind a solo project.

↗

Turso

SQLite-based, edge-distributed database. Good fit if you want a database instance per user or tenant instead of one big shared DB.

↗