How to export your data from Railway

Railway has no one-click "export all my data" button. For Postgres databases, the documented way to get your data out is a logical dump with pg_dump (a portable, SQL-level file you create yourself), with scheduled volume backups and point-in-time recovery as Railway-internal safety nets that only restore back into Railway.

Formats: SQL dump (pg_dump custom format, .dump), Plain SQL (pg_dump --format=plain)Plan: Any plan — pg_dump only needs the database connection string available on the Postgres serviceRole: Project member with access to the Postgres service's connection variables

✓ Verified against Railway's official documentation, October 2026

Step by step

1Open your Railway project and select the Postgres service.
2Install the Railway CLI and link your local project directory with `railway link`.
3Open a local tunnel to the database: run `railway connect postgres --tunnel-only`; this prints the host, port, user, password and database for a localhost connection.
4In a second terminal, run pg_dump against that local tunnel connection string with `--format=custom --no-owner --file=backup-$(date +%Y%m%d-%H%M%S).dump`.
5Alternative without the CLI tunnel: copy the DATABASE_PUBLIC_URL variable from the Postgres service and run pg_dump directly against it through Railway's public TCP proxy, using the same --format=custom --no-owner --file flags.
6Optional safety net: on the Postgres service, open the Backups tab and enable scheduled volume backups and/or point-in-time recovery — these restore inside Railway only, not to a downloadable file.
7Move the resulting .dump file off Railway's infrastructure (e.g., to your laptop or another storage provider) so you have an independent copy.

What the export includes

✓Full schema and data for the Postgres database at the time of the dump (a complete SQL-level copy)

What it doesn't include

✗Application source code, build/start commands, and project config — these live in your connected GitHub repo and Railway project settings, not in the database dump
✗Environment variables — these must be exported separately (e.g., via the Railway CLI) and are not part of a pg_dump
✗Other database engines on Railway (MySQL, Redis, MongoDB) — Railway's published backup/restore guide is Postgres-specific; other engines need their own native dump tool
✗A downloadable copy from volume backups or point-in-time recovery — both restore only to a new/same Postgres service inside Railway, not to a file you can take elsewhere

Before you start

⚠Custom-format dumps (--format=custom) are compressed and allow selective/parallel restores with pg_restore, unlike plain SQL dumps
⚠Point-in-time recovery restores to a brand-new sibling Postgres service, not in place over the original
⚠Volume backups restore only to the same service, project, and environment they came from
⚠You need either the Railway CLI with a tunnel, or the DATABASE_PUBLIC_URL variable, to reach the database from outside Railway — there's no UI button that packages the whole database for download

Where to take your data

Free and cheaper tools that can take a Railway export. See all Railway alternatives →

Render Free

Create a Postgres instance on Render, get its external connection string, and run pg_restore (or psql for plain-format dumps) against it using the .dump file produced from Railway.

Fly.io

Create a Postgres app on Fly, open a proxy with `fly proxy` or use its public connection string, then run pg_restore against it with the Railway-exported .dump file.

Railway Starter

Alternatives

If you're just downsizing plans rather than leaving, spin up a new Postgres service on Railway's free/Hobby trial and run pg_restore against its connection string using the same dump file.

Exporting from Railway: common questions

Does Railway have a built-in button to export my whole database?+

No. Railway's own guide describes the logical dump as something you create yourself with pg_dump, not an automated export feature.

What format is the exported file in?+

Railway's documented approach uses pg_dump with --format=custom, which produces a compressed binary dump that pg_restore can restore selectively or in parallel; plain SQL format is also an option with pg_dump.

Can I restore the dump somewhere other than Railway?+

Yes. Railway's guide explicitly lists logical dumps as restorable anywhere — another Railway service, another provider, or your own laptop.

Do I need the Railway CLI to get my data out?+

No. You can skip the CLI tunnel and run pg_dump directly against the DATABASE_PUBLIC_URL variable through Railway's public TCP proxy.

What's the difference between a volume backup and a pg_dump?+

Volume backups are scheduled/manual snapshots that restore only to the same Postgres service, project and environment, while a pg_dump is a portable SQL-level file you can restore anywhere.

I'm using MySQL/Redis/MongoDB on Railway — does this guide apply?+

Railway's published backup/restore documentation covers Postgres specifically; for other database templates you'd use that engine's own dump tool (e.g., mysqldump, redis-cli --rdb) over the same CLI tunnel approach.