Perimattic
ManageStacks · Databases

Managed pgvector Hosting — production-ready from $11 a month

Vector similarity search extension that turns PostgreSQL into an embeddings database. Deployed on your own dedicated instance in AWS, Azure, or GCP, kept patched, backed up, and monitored by ManageStacks — standard pgvector, no lock-in.

pgvector on ManageStacks is a PostgreSQL database with the open-source pgvector extension enabled, deployed to your own AWS, Azure, or GCP region — flat per-app pricing from $11 a month ($25 on the Standard production plan). It stores embeddings next to your regular tables, so AI teams can skip a separate vector service such as Pinecone.

Daily backups includedAWS · Azure · GCPData export any time24×7 SRE available
pgvector logo
pgvector
Vector similarity search extension that turns PostgreSQL into an embeddings database
The application

What does pgvector do, and why do teams deploy it?

pgvector adds a vector data type and similarity search to PostgreSQL. You store embeddings produced by a language or image model in an ordinary column, then query for the nearest neighbours using distance operators for L2, inner product, or cosine similarity. Because it runs inside Postgres, vectors live in the same transactions, backups, and access rules as the rest of your data, and you can join them with relational tables in a single SQL query. For larger collections it offers approximate indexes, including HNSW and IVFFlat, that trade a little recall for speed. It is written in C and loads as a standard extension.

  • Native vector column type for storing model embeddings in Postgres
  • Distance operators for L2, inner product, and cosine similarity
  • HNSW indexes for fast approximate nearest-neighbour search
  • IVFFlat indexes as a lighter-weight approximate option
  • Combine vector search with SQL filters and joins in one query
  • Exact nearest-neighbour search when you need perfect recall
Abstract visualization of structured data flowing between systems
Databases

Vector similarity search extension that turns PostgreSQL into an embeddings database

Open-source alternative

pgvector as a self-hosted alternative to Pinecone, Weaviate Cloud and Zilliz Cloud

Instead of paying per-seat fees to a SaaS vendor, teams self-host pgvector to keep full control of their data and infrastructure. ManageStacks removes the operational overhead — you get the cost and privacy benefits of open source with the reliability of a managed platform.

Replaces PineconeReplaces Weaviate CloudReplaces Zilliz Cloud
Pricing

What does managed pgvector hosting cost?

Flat per-app pricing, in your chosen AWS, Azure, or GCP region. No per-user pricing — a busy deployment costs the same as a quiet one.

Starter

$11/app/mo

Staging and internal tools. Dedicated instance, TLS, daily backups, managed upgrades.

Standard

$25/app/mo

Production workloads. Adds monitoring, staging environment, region choice, priority support.

Business

$45/app/mo

High-traffic and compliance workloads. Adds a high-availability replica and same-day support.

24×7 SRE retainer

$499/mo

Round-the-clock on-call across every hosted application, for teams that need a pager answered at 3am.

The honest answer

When is self-hosting pgvector the right answer instead?

“Managed hosting is not always the correct call.”

Self-host when a platform team already runs the infrastructure and on-call rotation to operate pgvector at genuinely low marginal cost. Self-host when compliance requires an air-gapped or on-premises deployment that no hosted option can satisfy. And self-host when the deployment depends on heavy customisation with a fast internal build-deploy loop, because an internal release process will beat any managed change process.

For everyone else — teams whose engineers have better things to do than shepherd upgrades — managed hosting is cheaper than the hours it replaces.

Infrastructure

Which cloud should pgvector run on — AWS, Azure or GCP?

For most workloads, the choice of cloud matters less than proximity: run pgvector in the same cloud and region as the applications and data it talks to, because every request between them adds a round trip. The underlying compute performs equivalently across AWS, Azure, and GCP.

In practice, an existing cloud footprint decides it. All plans support all three clouds, and moving regions later is a scheduled migration, not a rebuild.

AWS logo
AWS

Deepest managed-service catalog, default when there's no existing footprint

Azure logo
Azure

Best fit for teams already on Microsoft 365 or Entra ID

GCP logo
GCP

Strongest for data/analytics-adjacent workloads

Close-up of data centre server hardware and cabling
Multi-cloud

Every plan supports AWS, Azure, and GCP — region choice included.

FAQ

Common questions about pgvector on ManageStacks

What is pgvector on ManageStacks?

It is a PostgreSQL instance with the pgvector extension ready to use, running in your own cloud account. ManageStacks handles provisioning, SSL, and backups, so you can create vector columns and start storing embeddings straight away.

How does pgvector compare to Pinecone?

Pinecone is a dedicated, proprietary vector service. pgvector keeps embeddings inside Postgres, so you avoid syncing data between systems and can filter vectors with normal SQL. A purpose-built service may suit extreme scale, while pgvector suits teams that already rely on Postgres.

Which index types does pgvector support?

pgvector supports exact search by default and two approximate index types, HNSW and IVFFlat. HNSW generally gives better speed and recall trade-offs but takes longer to build, while IVFFlat builds quickly and uses less memory. You choose per column based on your workload.

Can I move an existing pgvector database to ManageStacks?

Yes. Since it is standard PostgreSQL, you can restore a dump or use logical replication from your current database. You keep full ownership of the data, and you can export it again at any time with ordinary Postgres tools.

How are pgvector backups and upgrades handled?

Backups follow the normal PostgreSQL approach, so your embeddings are protected along with every other table. Extension and Postgres upgrades are managed for you. After a major version change, it is wise to rebuild vector indexes and re-check query plans.

Run pgvector without carrying the pager

Subscribe through your AWS, Azure, or GCP marketplace. We handle provisioning, SSL, monitoring, backups, updates, and security. From $11/mo.