PostgreSQL vs MySQL: Choosing a Database

Both are excellent, battle-tested relational databases. If you're choosing in 2026, the answer is usually PostgreSQL — but let's be specific about why, and when MySQL still wins.

Where PostgreSQL wins

• Integrity: stricter by default — no silent truncation or loose modes.
• JSONB: indexable JSON with real querying, not a blob.
• Extensions: PostGIS, pgvector, TimescaleDB, full-text search.
• Concurrency: MVCC gives true concurrent reads without locking.

Where MySQL wins

• Massive ecosystem and hosting ubiquity (WordPress-era stacks).
• Simple replication setups with a huge body of operational knowledge.
• Slightly lower operational overhead for very simple workloads.

A telling difference

-- PostgreSQL: strict, explicit, type-safe
CREATE TABLE events (
    id bigserial PRIMARY KEY,
    payload jsonb NOT NULL,          -- indexable JSON
    happened_at timestamptz NOT NULL DEFAULT now()
);
CREATE INDEX ON events USING GIN (payload);

-- pgvector: store and search embeddings right in Postgres
CREATE EXTENSION vector;
CREATE TABLE docs (id bigserial PRIMARY KEY,
                   embedding vector(1536));
CREATE INDEX ON docs USING ivfflat (embedding vector_cosine_ops);

The recommendation

Default to PostgreSQL. It's the most capable open-source relational database, and modern AI-era features (pgvector for embeddings, JSONB for semi-structured data) make it the natural hub for most applications we build.

If you're extending an existing MySQL codebase or building on WordPress, stick with MySQL — migrating for its own sake is rarely worth it.

Whichever you choose, normalise your schema, index for your actual queries, and use transactions. The database is the part of the system you can't easily rewrite later.