Deploy and Host Meilisearch with Persistent Indexes on Railway

Search API with persistent indexes, production auth and a generated key.

About Hosting Meilisearch with Persistent Indexes

Deploying this template creates Meilisearch in your Railway project. The application receives an HTTPS domain automatically. Database and Redis services, when included, use private networking and generated credentials. Persistent data is stored on the included volumes or object storage, so application redeploys do not discard it.

Why Deploy Meilisearch with Persistent Indexes on Railway

Official Meilisearch image with production authentication enabled.
A unique generated master key and automatic HTTPS domain.
Indexes, dumps and daily local snapshots share one persistent volume.

Common Use Cases

Add fast search to a product catalog.
Search documentation or an internal knowledge base.
Provide a search backend for a web application.

Dependencies for Meilisearch with Persistent Indexes

A Railway account and a plan with capacity for 1 service(s).
Included services: Meilisearch.
1 persistent volume(s), provisioned by this template.
External integrations, mail delivery and your own business data are configured after installation as needed.

First launch

Deploy and open the generated service domain.
Read MEILI_MASTER_KEY from the service variables for server-side administration.
Create an index and import documents using a Meilisearch SDK or REST API.
Generate a scoped search-only key for your frontend.

Example

Set the URL and credential variables locally to your own deployment values.

curl --fail --header "Authorization: Bearer $MEILI_MASTER_KEY" "$MEILI_URL/keys"

Operations

Local snapshots are on the same volume as indexes; keep an off-volume backup as well. Before a version upgrade create a dump and follow the upstream migration procedure. Never put the master key in browser code. This single-node service does not provide replication or high availability.

Images are pinned to verified upstream image digests captured on 2026-10-17. Review upstream release notes, back up persistent data and update the image deliberately when upgrading. Railway charges the deployer for running services, storage and network usage.

Support and upstream

For Railway template wiring or deployment issues, use the discussion/support link on this template page. Include service names and redacted logs; omit passwords and tokens. Application documentation:

Configuration reference · Meilisearch source