Laravel Cloud — a full look at the Laravel-native PaaS

Laravel Cloud, launched in 2025. Learn how it differs from Forge, what it’s good and bad at, and the benefits of using it in practice.

What is Laravel Cloud?

Laravel Cloud is Laravel’s official fully managed PaaS. Everything from app deployment, scaling, databases, and caching to WebSocket infrastructure is integrated for Laravel.

The Cloud Changelog records Launched Laravel Cloud! on February 24, 2025. It’s a relatively new service that reached general availability in 2025.

---

How it differs from Forge

Laravel Cloud and Laravel Forge are both for Laravel developers, but operate at different layers.

Aspect	Laravel Cloud	Laravel Forge
Position	Fully managed PaaS	Server management tool
Primary operation	Choose environments and resources, then deploy	Provision and operate VPS / EC2, etc.
Infrastructure management	Minimal (managed by Cloud)	OS/middleware operation is your responsibility
Root access	None (no direct OS-level operations)	Yes (flexible control over servers)

If you want to “ship a Laravel app quickly without worrying about infrastructure,” Cloud is the right pick. If you want fine-grained server control, Forge is the better fit.

---

What Laravel Cloud excels at

Laravel Cloud’s strength is making it easy to enable the features you use in real-world Laravel operations from a dashboard-driven experience.

Task Scheduler

With Scheduled Tasks, simply enable Scheduler on an App cluster or Worker cluster to have schedule:run execute every minute.

Managed Queues

For Queues, Managed Queues are recommended. Laravel Cloud handles queue provisioning, worker startup, and scaling automatically. When there are no jobs, the worker count drops to zero, so you’re only billed while processing.

Option	Suitable use	Scaling	Billing model
Managed Queues	Production, preview, and development in general. The first choice	Auto-scales based on queue backlog; scales down to zero when idle	Worker uptime seconds + queue operations
Worker cluster	You want to manage queue:work yourself, or run heavier background workloads	CPU/memory-based	Standard Worker cluster pricing
App cluster background processes	Small-scale development or low-frequency jobs	CPU/memory-based	Included in App cluster

Managed Queues let you separate web traffic and job processing while also giving you a failed-jobs dashboard. On the other hand, if you want to manage a different driver (like database or redis) yourself, use a Worker cluster. If the same compute as the app is sufficient, use the App cluster’s background processes.

1. Create a Managed Queue

   Click New Managed Queue on the environment’s infrastructure canvas dashboard.

2. Choose the queue name and worker settings

   Configure queue name, instance size, polling interval, maximum workers, and so on.

3. Save and deploy

   Save the settings and redeploy—Laravel Cloud provisions the queue and worker pool automatically.

Because Managed Queues workers scale to zero when there are no jobs, they pair well with Scale to Zero (described below). Even if the App cluster is asleep, only queue processing can wake up automatically when needed.

Scale to Zero (Flex compute)

Flex, described under Compute, is a low-cost, flexible compute class. The new Scale to Zero evolves this Flex compute from the older Hibernation model. On the new Flex compute, HTTP-request wake-ups have been reduced to under 500ms—no more waiting 5–20 seconds as before.

To enable it, toggle Scale to Zero on in the App cluster settings, save, and redeploy.

Furthermore, in environments with Scale to Zero enabled, the environment automatically wakes to run Scheduled Tasks and process queues even while asleep. Running schedule:run on the App cluster while delegating job processing to Managed Queues is currently the most manageable combination, and matches what the official docs suggest.

Your own queue:work on the App cluster may be interrupted across sleep timeouts. If you need reliable queue processing, follow the official docs and prioritize Managed Queues.

MySQL Scale to Zero

The Flex size for Laravel MySQL also supports Scale to Zero. When a database receives no connections for a configured idle timeout (from 1 minute to 1 hour), it sleeps and compute billing stops. Storage stays online, so data is retained while the database is asleep and only storage billing continues.

The database wakes automatically when the next connection arrives. Every Laravel MySQL cluster is fronted by a proxy layer, so connections made during startup are held by the proxy and then forwarded to the database. This does not fail the connection; the first query is only delayed by a few hundred milliseconds (larger memory sizes tend to take longer to wake). Automatic backups and manual snapshots also wake the database temporarily, then it returns to sleep after completion.

At the moment, this is available for Flex sizes in US East (N. Virginia) and US East (Ohio), with other regions planned. Existing databases continue running at their current size and price. Pro sizes are always-on and are not eligible for Scale to Zero.

Scale to Zero Flex databases provide little benefit for databases shared by multiple apps or accessed throughout the day, because they will not sleep often. If an existing MySQL database is already stable, switching to the Scale to Zero-enabled version may slightly raise the minimum cost ($5.50 → $6.60), so decide based on the workload. Newly created MySQL databases already only offer Scale to Zero-enabled options.

Combining this with Compute Scale to Zero and Laravel Valkey Scale to Zero can minimize costs for development environments, side projects, and even production workloads with bursty traffic.

Laravel Octane

Enabling Use Octane as runtime in Compute has Cloud run the app with the Octane + FrankenPHP setup. There’s no need to piece together server configuration by hand.

Inertia SSR

You can enable Use Inertia SSR from Compute. When using a starter kit, npm run build:ssr is recommended.

WebSocket server (Reverb)

With Laravel Cloud WebSockets, you can create Reverb-based WebSocket clusters. Attaching one to an environment automatically injects REVERB_* and VITE_REVERB_* variables.

Laravel Nightwatch integration

Environments offers a one-click integration, and the Cloud Changelog records One-Click Nightwatch Integration on July 1, 2025. If you need to configure it manually, see nightwatch-on-cloud for instructions.

---

Pricing and free plan

Laravel Cloud pricing is essentially plan fee + usage-based billing. Always check Pricing for the details.

• Starter: $5/month. Includes $5 of usage credit each month. New signups get the first month free
• Growth: $20/mo + usage, with $5 of usage credit each month
• Business: $200/mo + usage, with $5 of usage credit each month

Starting a small project on Starter is now easier than before. Combining Scale to Zero with Managed Queues makes costs more predictable, even for indie projects or internal tools with long idle periods.

---

Things to keep in mind in practice

You can’t rely on root-privileged operations

Because Cloud is a PaaS, it prioritizes operational simplicity over OS-level freedom. If you need work that assumes root privileges—kernel tuning, running custom daemons—consider options like Forge.

Control Nightwatch event volume up front

When using Nightwatch on a free tier or small setup, event volume can grow quickly, so it’s safer to configure sampling from the start.

NIGHTWATCH_REQUEST_SAMPLE_RATE=0.1
NIGHTWATCH_IGNORE_QUERIES=true

---

Good fits and poor fits

Good fits

• You want to ship a Laravel app to production as fast as possible
• You want to run Scheduler / Queue / Octane / Inertia SSR / Reverb together
• You want to focus on app development and reduce server operations cost

Poor fits

• You want strong OS-level tuning with root privileges
• You have strong existing custom infrastructure needs (proprietary middleware, unusual network constraints)

---

Summary

Laravel Cloud is a strong fit for teams that want to “run Laravel’s advanced features with less infrastructure overhead.” Projects combining Scheduler / Queue / Octane / Inertia SSR / Reverb / Nightwatch in particular can start up and iterate much faster.

If you need root-privileged operations, however, Forge is a better fit. The safest path is to run your actual workload on the free trial first and check whether it matches your operational requirements.