---
title: "Deploy Hatchet Lite"
description: "Durable task queue and workflow engine for background jobs and AI agents"
category: "Queues"
url: https://railway.com/deploy/7oEFNd
---

# Deploy Hatchet Lite

Durable task queue and workflow engine for background jobs and AI agents

**[Deploy Hatchet Lite on Railway](https://railway.com/template/7oEFNd)**

Machine-readable deploy manifest (JSON, validated by TemplateCI): https://railway.com/deploy/7oEFNd/manifest.json

- **Creator:** withSeismic
- **Category:** Queues
- **Total deploys:** 31

## Template content

### Postgres https://devicons.railway.app/i/postgresql.svg

- **Image:** ghcr.io/railwayapp-templates/postgres-ssl:18

### hatchet-lite https://avatars.githubusercontent.com/u/118301880?.png

- **Image:** ghcr.io/hatchet-dev/hatchet/hatchet-lite:latest
- **Health check:** /api/ready
- **Public domain:** Yes

## Documentation

# Deploy and Host Hatchet on Railway

Hatchet is an open-source (MIT) platform for background tasks and durable workflows, built on Postgres. Workers written with the Python, TypeScript, Go or Ruby SDKs pick up tasks with retries, timeouts, concurrency and rate limits, cron schedules and event triggers, while a dashboard shows every run with its logs, inputs and outputs. It is a self-hosted alternative to Celery, BullMQ, Temporal, Inngest and Trigger.dev.

## About Hosting Hatchet

This template deploys the latest Hatchet Lite, the official single-container edition with the engine, REST API and dashboard, backed by Railway Postgres 18. It works out of the box and is safe to leave on the public internet:

- **No default login.** The admin account is created from `ADMIN_EMAIL` and a randomly generated `ADMIN_PASSWORD` (see the hatchet-lite service's Variables tab). Public sign-up is turned off.
- **Keys survive redeploys.** Hatchet's generated encryption keys and config live on a Railway volume at `/config`, so API tokens keep working after every restart and upgrade.
- **Workers can connect from anywhere.** A Railway TCP proxy publishes Hatchet's gRPC port (7077), and every API token you create already carries that address.
- **Postgres only.** Queues, run history and pub/sub all run on Postgres, so there is no RabbitMQ or Redis to pay for.

### Getting started

1. Deploy, then open the hatchet-lite service's public URL.
2. Log in with your `ADMIN_EMAIL` and the generated `ADMIN_PASSWORD` from the Variables tab.
3. Go to **Settings → API Tokens** and create a token.
4. Start a worker with two environment variables:

```bash
HATCHET_CLIENT_TOKEN=
HATCHET_CLIENT_TLS_STRATEGY=none
```

The worker appears on the Workers page within seconds. Trigger a task from your code or the dashboard and the run turns green.

**Workers on Railway:** to keep worker traffic on Railway's private network, also set `HATCHET_CLIENT_HOST_PORT=hatchet-lite.railway.internal:7077` on the worker service. This is recommended for production because the public TCP proxy carries gRPC without TLS (workers still authenticate with their token).

## Common Use Cases

- Background jobs and queues for web apps, with retries, timeouts and full visibility
- Durable multi-step workflows, AI agent loops and LLM pipelines that survive restarts
- Cron jobs, scheduled tasks and event-driven processing with concurrency and rate limits
- Fan-out data ingestion, document processing and batch work

## Dependencies for Hatchet Hosting

- `ghcr.io/hatchet-dev/hatchet/hatchet-lite:latest` (official Hatchet image)
- Railway Postgres 18 (`ghcr.io/railwayapp-templates/postgres-ssl:18`) with a volume
- A Railway volume at `/config` for Hatchet's keys and generated config
- A Railway TCP proxy on port 7077 for worker gRPC connections

### Deployment Dependencies

- [Hatchet documentation](https://docs.hatchet.run/)
- [Hatchet Lite self-hosting guide](https://docs.hatchet.run/self-hosting/hatchet-lite)
- [Hatchet configuration options](https://docs.hatchet.run/self-hosting/configuration-options)
- [Hatchet on GitHub](https://github.com/hatchet-dev/hatchet)
- [Railway TCP proxy](https://docs.railway.com/reference/tcp-proxy)

### Implementation Details

Key variables on the hatchet-lite service:

| Variable | Value | Why |
|---|---|---|
| `SERVER_URL` | `https://${{RAILWAY_PUBLIC_DOMAIN}}` | Dashboard and API address embedded in tokens |
| `SERVER_GRPC_BROADCAST_ADDRESS` | `${{RAILWAY_TCP_PROXY_DOMAIN}}:${{RAILWAY_TCP_PROXY_PORT}}` | Public gRPC address workers dial |
| `SERVER_INTERNAL_CLIENT_INTERNAL_GRPC_BROADCAST_ADDRESS` | `localhost:7077` | Keeps Hatchet's internal client inside the container |
| `SERVER_ALLOW_SIGNUP` | `false` | Only the seeded admin can log in |
| `SERVER_MSGQUEUE_KIND` | `postgres` | No message broker needed |

The image tracks `latest`, so every new deploy starts on the current Hatchet release, including Hatchet v2 features (durable streams, operators) as they ship in the Lite image. To upgrade an existing install, redeploy the hatchet-lite service; migrations run automatically on boot. To hold a version, set the image tag to a specific [release](https://github.com/hatchet-dev/hatchet/releases). Back up Postgres regularly, because it holds every queue and run.

## Why Deploy Hatchet 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 Hatchet 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.


## Similar templates

- [Celery | Web, Worker and Scheduler as Three Services](https://railway.com/deploy/celery-or-web-work-1) — Web, worker and scheduler wired up: Celery on Redis, outcomes in Postgres.
- [Redpanda](https://railway.com/deploy/redpanda-1) — Redpanda 26.2: Kafka-compatible streaming broker with Kafbat UI console
- [smoothmq](https://railway.com/deploy/AJv-64) — A drop-in replacement for AWS SQS

Open this page in a browser: https://railway.com/deploy/7oEFNd
