Pyyan / Compare / turbopuffer vs pgvector
turbopuffer vs pgvector
2 of 5Vector Databases · verified 13 Aug 2026
turbopufferturbopuffercurrentpgvectorPostgreSQLcurrent
| Specification | turbopuffer | pgvector |
|---|---|---|
| Summary | Vectors on object storage, priced accordingly. | Vectors inside the database you already run. |
| Index | Proprietary, object storage | HNSW, IVFFlat |
| Hybrid search | Yes | Yes, with SQL |
| Hosting | Managed only | Self-host or any managed Postgres |
| Licence | Proprietary | PostgreSQL |
| p50 latency | ~20ms | ~15ms |
| Category | Vector Databases | Vector Databases |
| Official | turbopuffer ↗ | PostgreSQL ↗ |
Highlighted rows are where these differ.
turbopuffer
- Built on object storage rather than RAM
- Notion cut search costs ~60% moving from Pinecone Serverless
Best for large corpora where cost dominates.
pgvector
- One database for vectors, rows and joins
- The honest default: reach for something else only when this stops working
Best for almost everyone, until scale says otherwise.