Deploy and Host PostgreSQL with pgvectorscale and pg_textsearch with Railway

PostgreSQL 18 with pgvector, pgvectorscale, and pg_textsearch, for vector, BM25 keyword, and hybrid search in one database. pgvectorscale adds StreamingDiskANN indexes for large embedding tables, and pg_textsearch adds BM25 ranking. All three extensions are created for you in the default database.

About Hosting PostgreSQL with pgvectorscale and pg_textsearch

The template runs one PostgreSQL service from a tested image for amd64 and arm64, with a volume at /var/lib/postgresql/data and a generated password. The image builds on Railway's own postgres-ssl image, so it keeps SSL, pgBackRest WAL archiving, point-in-time recovery, and Railway's volume checks. pg_textsearch has to be preloaded, and the image keeps shared_preload_libraries correct on every boot. It also creates missing extensions and updates their SQL to match the image, so an image update never leaves a database on an old extension version. Patch releases install automatically during Railway's nightly update window.

Common Use Cases

Hybrid search that ranks by both meaning and keywords
Retrieval-augmented generation (RAG) with BM25 and vector retrieval
Full-text search with BM25 ranking instead of ts_rank
Storing embeddings and documents next to the relational data they describe

Dependencies for PostgreSQL with pgvectorscale and pg_textsearch Hosting

None. Applications connect with DATABASE_URL.

Deployment Dependencies

Implementation Details

Connect other services with the DATABASE_URL reference variable. It uses private networking, and DATABASE_PUBLIC_URL goes through the TCP proxy.

CREATE TABLE documents (
    id BIGSERIAL PRIMARY KEY,
    content TEXT,
    embedding VECTOR(1536)
);

CREATE INDEX ON documents USING diskann (embedding vector_cosine_ops);
CREATE INDEX ON documents USING bm25(content) WITH (text_config='english');

-- BM25: &lt;@&gt; returns the negative score, so lower is better.
SELECT id FROM documents ORDER BY content &lt;@&gt; 'search query' LIMIT 10;

-- Hybrid
SELECT id FROM documents
ORDER BY 0.7 * (embedding &lt;=&gt; $1) + 0.3 * (content &lt;@&gt; $2)
LIMIT 10;

Why Deploy PostgreSQL with pgvectorscale and pg_textsearch on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying PostgreSQL with pgvectorscale and pg_textsearch on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.