Relational (SQL): tables, schemas and joins with strong consistency. PostgreSQL, MySQL.
Document: flexible JSON-like documents. MongoDB. Key-value: fast lookups by key, often in memory. Redis. Wide-column: huge write-heavy datasets. Cassandra. Graph: relationships first. Neo4j. Time-series and vector databases serve metrics and AI search.
Why it matters: pick by access pattern, not hype. Every type has its own injection risk too: SQL injection has NoSQL cousins.