Skip to main content

What is Laravel Cloud

Laravel Cloud is Laravel’s official fully managed PaaS.
It integrates deployment, scaling, databases, cache, and WebSocket infrastructure for Laravel applications.
The Cloud changelog records Launched Laravel Cloud! on February 24, 2025, so it is a relatively new GA service.

How it differs from Forge

Laravel Cloud and Laravel Forge both target Laravel developers, but they solve different layers of operations.
PerspectiveLaravel CloudLaravel Forge
PositioningFully managed PaaSServer management tool
Primary workflowChoose environment/resources and deployProvision and manage VPS / EC2 and operate servers
Infrastructure responsibilityMinimal (managed by Cloud)OS and middleware operations are your responsibility
Root accessNot providedProvided
Choose Cloud when you want to ship Laravel apps fast without managing infrastructure internals.
Choose Forge when you need detailed server-level control.

Where Laravel Cloud is strong

Laravel Cloud makes common Laravel production features easy to enable from the dashboard.

Task scheduler

With Scheduled Tasks, enabling Scheduler on an app or worker cluster runs schedule:run every minute.

Managed Queues

In Queues, Managed Queues are now the recommended default. Laravel Cloud handles queue provisioning, worker startup, and scaling automatically. When no jobs are pending, workers can scale down to zero so you are billed only while jobs are processed.
OptionBest forScaling modelCost model
Managed QueuesMost production, preview, and development environmentsAutomatic based on queue backlog. Can scale to zero when idleWorker runtime seconds + queue operations
Worker clusterLarger background workloads where you want to manage queue:work directlyCPU / memory-basedStandard Worker cluster pricing
App cluster background processesSmall development use cases and low-frequency jobsCPU / memory-basedIncluded in App cluster compute
Managed Queues separate web traffic from job processing and provide failed job visibility in the dashboard.
If you need to self-manage drivers like database or redis, use a Worker cluster. If sharing compute with your app is enough, use App cluster background processes.
1

Create a Managed Queue

In your environment’s infrastructure canvas dashboard, click New Managed Queue.
2

Configure queue and worker settings

Set queue name, instance size, polling interval, and max workers.
3

Save and redeploy

Save settings and redeploy. Laravel Cloud provisions the queue and worker pool automatically.
Managed Queues can scale workers down to zero when no jobs exist, which pairs well with Scale to Zero. Even if your App cluster sleeps, queue processing can still auto-start when required.

Scale to Zero (Flex compute)

Compute explains that Flex is the low-cost, flexible compute class. The new Scale to Zero feature evolved from legacy hibernation. On Flex compute, cold starts on HTTP requests are now under 500ms, instead of the older 5–20 second behavior. Enable it by turning on Scale to Zero in your App cluster settings, then saving and redeploying. When Scale to Zero is enabled, the environment can also auto-start while sleeping for Scheduled Tasks and queue processing.
Running schedule:run on the App cluster while delegating jobs to Managed Queues is now the most practical default architecture recommended in Laravel Cloud’s official documentation.
A self-managed queue:work process on the App cluster can still be terminated when the cluster sleeps. If reliable queue continuity matters, prefer Managed Queues as documented.

Laravel Octane

In Compute, enabling Use Octane as runtime runs your app with Octane + FrankenPHP without manual server setup.

Inertia SSR

In Compute, enabling Use Inertia SSR makes SSR setup straightforward.
When using starter kits, npm run build:ssr is recommended.

WebSockets (Reverb)

With Laravel Cloud WebSockets, you can create a Reverb-based WebSocket cluster.
After attaching it to an environment, REVERB_* and VITE_REVERB_* variables are injected automatically.

Laravel Nightwatch integration

Environments provides one-click integration, and the Cloud changelog records One-Click Nightwatch Integration on July 1, 2025.
If you need manual setup, see nightwatch-on-cloud.

Pricing and free trial

Laravel Cloud pricing is generally plan fee + usage.
Always verify current details in Pricing.
  • Starter: $5/mo, includes $5 monthly usage credits, and first month free for new signups
  • Growth: $20/mo + usage, includes $5 monthly usage credits
  • Business: $200/mo + usage, includes $5 monthly usage credits
Free trial includes $5 in credits with no credit card required, so you can test on real infrastructure. Starter is now easier to adopt for small projects.
Combining Scale to Zero with Managed Queues makes costs more predictable for personal projects and internal tools with long idle periods.

Important limitations

No root-level server operations

Because Cloud is a PaaS, it prioritizes operational simplicity over OS-level control.
If your workload requires kernel tuning, custom daemons, or other root-based operations, Forge or another server-centric option is usually a better fit.

Control Nightwatch event volume early

If you use Nightwatch on smaller plans, event volume can grow quickly. Start with sampling and filtering settings.
NIGHTWATCH_REQUEST_SAMPLE_RATE=0.1
NIGHTWATCH_IGNORE_QUERIES=true

Good fit vs. poor fit

Good fit

  • You want to launch Laravel apps quickly.
  • You want to operate Scheduler / Queue / Octane / Inertia SSR / Reverb together.
  • You want to reduce infrastructure operations overhead and focus on application delivery.

Poor fit

  • You need heavy OS-level tuning with root access.
  • You have strict custom infrastructure constraints (special middleware, network restrictions, or similar).

Summary

Laravel Cloud is a strong option when your team wants to run advanced Laravel workloads with less infrastructure burden.
It is especially effective when your project combines Scheduler, Queue workers, Octane, Inertia SSR, Reverb, and Nightwatch.
If root-level control is a hard requirement, Forge remains the better option.
The safest path is to validate your actual workload with the free trial and decide based on operational requirements.
Last modified on June 3, 2026