{"slug": "pg-pinecone-now-deploys-on-railway-in-one-click", "title": "pg_pinecone now deploys on Railway in one click", "summary": "Pinecone released a one-click Railway template that deploys Postgres 18 with its open-source pg_pinecone extension installed, preloaded and enabled, the company said. In Pinecone benchmarks on 1 million vectors, the pg_pinecone index was about 15 times smaller than a pgvector HNSW index using halfvec and returned results faster at the same recall, which matters because Railway bills for the memory and storage a service uses. The pg_pinecone extension is available under Apache 2.0, and its default native engine runs inside Postgres without a Pinecone account; a Pinecone API key and one SQL call can move a table's search to a serverless index in Pinecone's managed service and back.", "body_md": "With pgvector, the HNSW graph index needs more memory as a table grows, which makes vector search in Postgres increasingly expensive to run. The [pg_pinecone Postgres extension](https://www.pinecone.io/blog/introducing-pg-pinecone/) has lower latency than pgvector at the same recall and uses less memory, so it costs less to run as tables grow. The new [Railway template](https://railway.com/deploy/pg-pinecone) for pg_pinecone gives developers Postgres with Pinecone's vector search built in, deployed in one click.\n\nThe Railway template runs Postgres 18 with pg_pinecone installed, preloaded, and enabled, so there's no extension to install or image to build. In our benchmarks on 1 million vectors, the pg_pinecone index was about 15 times smaller than a pgvector HNSW index using halfvec and returned results faster at the same recall. Railway bills for the memory and storage a service uses, so a smaller index lets the same Postgres service hold many more vectors for the same cost. When a table needs to scale, a Pinecone API key and one SQL call [move its search](https://docs.pinecone.io/guides/pg-pinecone/switch-to-remote) to a serverless index in Pinecone's managed service, and another call moves it back. Wherever the vectors are stored, queries don't change.\n\nThe pg_pinecone template is available now in Railway's template marketplace, and the pg_pinecone extension is [open source](https://github.com/pinecone-io/pg_pinecone) under Apache 2.0. The `native` engine, pg_pinecone's default, runs inside Postgres and needs no Pinecone account.\n\nTo get started, deploy the [template](https://railway.com/deploy/pg-pinecone) on Railway, or read the [pg_pinecone quickstart](https://docs.pinecone.io/guides/pg-pinecone/quickstart#hosting-platform).\n\nWas this article helpful?", "url": "https://wpnews.pro/news/pg-pinecone-now-deploys-on-railway-in-one-click", "canonical_source": "https://www.pinecone.io/blog/pg-pinecone-railway/", "published_at": "2026-11-18 16:00:00+00:00", "updated_at": "2026-11-18 16:12:28.820944+00:00", "lang": "en", "topics": ["ai-infrastructure", "developer-tools", "ai-tools"], "entities": ["Pinecone", "Railway", "pg_pinecone", "Postgres", "pgvector", "Apache 2.0"], "also_reported_by": [], "takeaway": "Deploying the Railway template gives Postgres 18 with pg_pinecone preloaded, and on 1 million vectors its index was about 15 times smaller than a pgvector HNSW halfvec index at equal recall and faster, with one SQL call plus a Pinecone API key moving search to a serverless index.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/pg-pinecone-now-deploys-on-railway-in-one-click", "markdown": "https://wpnews.pro/news/pg-pinecone-now-deploys-on-railway-in-one-click.md", "text": "https://wpnews.pro/news/pg-pinecone-now-deploys-on-railway-in-one-click.txt", "jsonld": "https://wpnews.pro/news/pg-pinecone-now-deploys-on-railway-in-one-click.jsonld"}}