Blog FAQs Docs Contact EN
Log in Sign up

Deploy Supabase Vector on Antyxsoft Cloud

A PostgreSQL vector store for embeddings and semantic search — the database your RAG pipeline needs, in your own region.

By
Supabase
Licence
Open source — no additional charge
Deploy this app

Create an instance, pick Marketplace App at the template step, and choose it from the list. You pay for the instance.

Deploy Supabase Vector See instance pricing

What Supabase Vector gives you

Supabase Vector is PostgreSQL with pgvector and the tooling around it: a vector column type, indexes for approximate nearest-neighbour search, and the ordinary SQL you already know for everything else.

That combination is why it has become a default for retrieval-augmented generation — embeddings live in the same database as the rows they describe, so a semantic search and a relational filter are one query rather than two systems.

Key features

  • Vector search. Nearest-neighbour queries over embedding columns with indexes.
  • SQL alongside. Filter by tenant, date or permission in the same query as the search.
  • Postgres underneath. Backups, replication and extensions all behave normally.
  • Any embedding model. Store vectors from whichever model you use.
  • Self-hosted. Documents and embeddings stay inside your region.

What is in the image

ApplicationPostgreSQL with pgvector, preconfigured
ConfigurationTuned defaults; full access to configuration
AccessFull root access over SSH; network access per firewall policy
Suggested sizeFrom 4 vCPU / 8 GB RAM; add block storage for index growth

Good fit for

  • RAG pipelines Retrieval for an assistant grounded in your own documents.
  • Semantic search Meaning-based search over support content or a knowledge base.
  • Regulated AI Embeddings of sensitive documents kept in the EU.

Getting started

Four steps from sign-in to a running application. The full walkthrough, with screenshots, is in the knowledge base.

  1. 01
    Log in to the Cloud Portal

    Sign in at portal.antyxsoft.io and start creating an instance, choosing the region closest to your users.

  2. 02
    At Select Template, open Marketplace App

    In the instance creation flow, switch from operating-system images to the Marketplace App tab.

  3. 03
    Pick the template and deploy

    Select the ready-to-go template, confirm the plan and region, then review and deploy. The instance boots with the application installed.

  4. 04
    Finish the application setup

    Set the database password, restrict access to your private network, then create the vector table and index for your embedding dimension.

Before it carries real traffic

  • Attach a Cloud Firewall that opens only the ports this application needs, with management access limited to your own addresses.
  • Schedule backups covering both the database and the application's data directory.
  • Add a Block Storage volume for data that will outgrow the instance disk.
  • Put it inside a VPC when it needs private access to other services.

Support

Antyxsoft supports the image and the infrastructure it runs on. Application-level questions are also covered by the upstream project's own documentation and community.

Supported byAntyxsoft
Support hours08:00 – 17:00 EEST

What is covered

Antyxsoft covers

The image itself, deployment problems, the instance, storage, network and platform availability.

You control

Configuration, add-ons, content, application updates and anything you change inside the guest.

Need it managed?

Ask about managed operations — updates, monitoring and backup handled for you.

Frequently asked questions

How is this different from plain PostgreSQL?
It is PostgreSQL with pgvector and the surrounding tooling preconfigured — a vector column type and approximate nearest-neighbour indexes ready to use, instead of a manual extension build.
Which embedding models does it work with?
Any of them. You store the vectors your model produces; the database only needs the dimension to be consistent within a table.
Why host embeddings myself?
Because embeddings of your documents are your documents, statistically speaking. Self-hosting keeps them — and the source text — inside your chosen EU region.
All FAQs →

More in this category

Databases

ClusterControl

Run database clusters properly: deployment, monitoring, backups and failover from a single console.

Learn more → Databases

Elasticsearch

Add real search to your product, or index your logs, on a self-hosted Elasticsearch instance.

Learn more → Databases

MariaDB

Run a MySQL-compatible database on your own instance, tuned and ready for the applications that expect it.

Learn more → Databases

MongoDB

Run MongoDB yourself — flexible documents, aggregation pipelines and no per-operation billing.

Learn more → Databases

PostgreSQL

Run PostgreSQL on your own instance — full configuration control, no per-IOPS surprises, unmetered traffic.

Learn more →

Keep the embeddings in Europe

Deploy Supabase Vector in your region and run retrieval over your own documents without shipping them elsewhere.

Deploy Supabase Vector Browse all apps