OpenCalled 2022 · due 2022–2024 · Technology
Rise of Multi-Model Databases
The database landscape will shift toward multi-model engines that handle documents, graphs, key-value, and relational queries in a single system, reducing operational overhead and data duplication.
ShareLinkedIn
Evidence so far
- PostgreSQL reached #1 most-used database globally (~45% of developers, Stack Overflow 2025), driven heavily by multi-model capabilities — pgvector, JSONB, PostGIS, TimescaleDB.
- Oracle 23c launched as a 'converged' multi-model database (row, columnar, JSON, spatial, vector). AWS Neptune, Azure Cosmos DB, and MongoDB Atlas all now market primarily as multi-model platforms.
- Elastic, Couchbase, and MongoDB shipped graph, vector, or columnar engines inside their primary products in 2024, validating the convergence thesis.
- Startups increasingly default to Postgres + extensions over polyglot stacks, but no single 'winner' has emerged as the definitive multi-model standard — the pattern is converging across many engines rather than consolidating into one.