Every figure below was read off laravel.com/cloud/pricing and the Laravel Cloud pricing docs on 21 August 2026. Laravel Cloud charges a plan fee plus metered usage, so the sticker price of $5 a month is only the first line of the invoice. This page shows the published rates, the arithmetic that turns them into a monthly bill, and the point where running your own server costs less.
# free plan · no credit card required
Provision on Hetzner, DigitalOcean, AWS, Vultr, Linode, OVHcloud or any VPS you already own
Laravel Cloud pricing has two parts: a plan fee of $5 a month for Starter, $20 for Growth or $200 for Business, plus metered usage on top. Every paid plan includes $5 of monthly usage credit, and the first month on Starter is free. Compute bills per second only while your app is awake, from $0.00893 an hour for Flex 512 MB to $0.40476 an hour for Pro 8 vCPU, so an always awake Flex 1 GB app costs roughly $13 a month in compute alone. Serverless Postgres adds $0.003 per CPU hour plus $0.15 per GB month of storage. An app that sleeps most of the time can genuinely land at the $5 plan fee and nothing more.
Last updated August 2026
Starter is $5 a month, Growth is $20 and Business is $200. Enterprise is quoted. That fee buys features, not capacity: Growth unlocks Pro compute, preview environments, worker clusters and autoscaling to 10x replicas, while Business adds scheduled autoscaling, custom roles and an advanced WAF. Nothing about the plan fee covers the resources your app burns. Every plan except Enterprise includes $5 of usage credit, which cancels the Starter fee for a genuinely idle app.
Compute bills per second while the app is awake and stops when it sleeps. Laravel quotes a wake time under 500ms. This is the whole pricing model in one sentence: a hobby app used one evening a week bills a handful of hours, while a store with traffic at 3am bills all 730 hours in the month. Before you compare Laravel Cloud to anything, write down honestly how many hours a month your app has to be awake.
Flex scales to zero and comes in 512 MB, 1 GB and 2 GB. Pro is always on and runs from 1 vCPU with 4 GB up to 8 vCPU with 32 GB. Dedicated sits above that. The jump from Flex 2 GB at $0.03571 an hour to Pro 2 vCPU at $0.10119 an hour is roughly triple, and Pro does not sleep, so moving to Pro changes both the rate and the number of billed hours at the same time.
The managed database is Serverless Postgres 17 and it bills on its own meter: $0.003 per CPU hour, $0.15 per GB month for storage on the root branch, and another $0.15 per GB month for point in time recovery history. Database compute also sleeps, with its own configurable sleep timer. On small apps the database is a rounding error. On a data heavy app with long retention it stops being one.
Queue workers and schedulers are compute instances, so a worker that runs around the clock bills like a second always on app. Managed queue operations are $1 per million, counting dispatch, start, completion, visibility extensions and polling. Valkey cache is priced by size and class with monthly caps. Pro queues keep at least one worker running, which means a Pro queue has a floor cost even when the queue is empty.
The published rates differ by region. US East is typically the cheapest and the Asia Pacific and EU regions run higher for the same size. Laravel Cloud lists US East Virginia, US East Ohio, EU Central Frankfurt, EU West Ireland, EU West London, Canada Central, Asia Pacific Tokyo, Asia Pacific Singapore and Asia Pacific Sydney. Every dollar figure on this page is the US East rate, so check your own region before you budget.
A month is 730 hours. Decide how many of them your app, your workers and your database genuinely have to be awake. This single number decides whether Laravel Cloud is nearly free or the most expensive option on your shortlist.
Take the rate for the class and size you need, multiply by awake hours, and add the database CPU hours and storage. That is your usage. Do it for the app, then again for every worker that runs on its own schedule.
Your bill is the plan fee plus usage, less the $5 monthly credit. Because the credit is small, it only matters for apps that are asleep most of the month. Set a spending limit at the number you just calculated so an unexpected traffic spike pauses compute instead of the invoice.
Put the equivalent RAM and vCPU on a droplet or a Hetzner box, add a flat fee panel, and compare. Below roughly 200 awake hours a month Laravel Cloud usually wins. Above always on production, a fixed price server normally wins by a wide margin.
These are the published plan features and fees, not our estimates. The one number that is not on this table is what your app will actually burn in usage, because that depends entirely on awake hours and compute size. Enterprise is quoted per customer, so where a cell says custom, Laravel does not publish a figure and neither will we.
swipe to see all columns →
| Starter | Growth | Business | Enterprise | |
|---|---|---|---|---|
| Plan fee per month | $5 plus usage | $20 plus usage | $200 plus usage | Custom |
| Included usage credit | $5 | $5 | $5 | None listed |
| Trial | First month free | None | None | None |
| Compute class | Flex, scales to zero | Pro, always on | Largest sizes | Dedicated compute |
| Autoscaling | None | Up to 10x replicas | Unlimited and scheduled | Dedicated |
| Managed queues per environment | 1 | 10 | Unlimited | Unlimited |
| Preview environments | No | Yes | Yes | Yes |
| Worker clusters | No | Yes | Yes | Yes |
| Web application firewall | No | Basic | Advanced | Advanced |
| Team roles and permissions | No | Yes | Custom roles | Custom roles |
| Custom domains | 10 | 50 | 250 | 250 plus |
| Log retention | 1 day | 7 days | 30 days | 30 days |
| Private networking | No | No | Add on | Included |
| Root access to the machine | No | No | No | No |
This is the case Laravel Cloud is built for and it wins it outright. An internal tool used a few hours a week on Flex 512 MB bills cents of compute, the $5 credit absorbs it, and you pay the plan fee. No server you rent by the month can compete with a bill that goes to zero when nobody is using the app.
Once the app never sleeps, the meter runs 730 hours a month and the scale to zero advantage disappears. A Pro 2 vCPU instance with 8 GB works out near $74 a month in compute before the plan fee, the database or a single worker. At that point a fixed price server with the same specs is usually less than half.
Every app carries its own compute and database meter, so cost grows roughly linearly with the client list. There is no per server consolidation to exploit, because there is no server. Agencies that host many small client sites on shared capacity generally find a panel over their own boxes cheaper, and they keep the ability to put ten client sites on one machine.
Laravel Cloud is a managed platform, so there is no root shell, no per domain mailbox management and no installing a system package the platform does not offer. If your app needs a binary, a cron running arbitrary shell, or mailboxes on the client's own domain, price the platform out early rather than discovering the limit after you migrate.
Laravel Cloud costs the plan fee plus whatever your app burns. The plan fee is $5 a month on Starter, $20 on Growth and $200 on Business, and Enterprise is quoted. Every one of those plans includes $5 of monthly usage credit, and the first month on Starter is free. Usage is billed per second while resources are awake.
That structure makes a single sticker price impossible, so here is the arithmetic instead. The four scenarios below use Laravel's own published US East rates, 730 hours in a month, and nothing else. They are our calculation from Laravel's rate card, not prices Laravel quotes as packages.
| Scenario | Resources | Usage | Plan | Monthly total |
|---|---|---|---|---|
| Internal tool, awake 4 hours a month | Flex 512 MB, tiny Postgres | about $0.12 | Starter $5 | $5.00 |
| Small app, awake around the clock | Flex 1 GB, Postgres 0.5 CPU, 5 GB | $14.89 | Starter $5 | $14.89 |
| Production app, always on | Pro 1 vCPU 4 GB, Postgres 1 CPU, 20 GB | $39.95 | Growth $20 | $54.95 |
| Busy app with a worker | Pro 2 vCPU 8 GB, Flex 1 GB worker, Postgres 2 CPU, 50 GB, 5M queue ops | $103.79 | Growth $20 | $118.79 |
Read the first and last rows together, because they are the whole story. The same platform bills $5 and $119 depending on one variable: how many of the month's 730 hours your app has to be awake. The credit is subtracted from usage, so the totals above already account for it.
Laravel also publishes an interactive Laravel Cloud pricing calculator with four scenario presets. Use it to sanity check your own numbers, but do the multiplication yourself for the worker and database too, because those are the lines people forget.
Compute is quoted per worker second in the Laravel Cloud docs. Converted to an hourly rate and then to a full month at 730 awake hours, the US East rate card looks like this.
| Class and size | Per hour | 730 hours | Sleeps when idle |
|---|---|---|---|
| Flex 512 MB | $0.00893 | $6.52 | Yes |
| Flex 1 GB | $0.01786 | $13.04 | Yes |
| Flex 2 GB | $0.03571 | $26.07 | Yes |
| Pro 1 vCPU, 4 GB | $0.04762 | $34.76 | No |
| Pro 2 vCPU, 8 GB | $0.10119 | $73.87 | No |
| Pro 4 vCPU, 16 GB | $0.20238 | $147.74 | No |
| Pro 8 vCPU, 32 GB | $0.40476 | $295.47 | No |
| Dedicated 1 vCPU, 2 GB | $0.05989 | $43.72 | No |
| Dedicated 2 vCPU, 4 GB | $0.11938 | $87.15 | No |
| Dedicated 4 vCPU, 8 GB | $0.23876 | $174.29 | No |
| Dedicated 8 vCPU, 16 GB | $0.47751 | $348.58 | No |
Two things in that table matter more than the rest. First, Flex is the only class that scales to zero, which is why the difference between Flex and Pro is much bigger than the rate difference suggests: moving to Pro raises the rate and removes the discount you were getting from sleeping. Second, the Dedicated 1 vCPU with 2 GB costs more per hour than the Pro 1 vCPU with 4 GB, because you are paying for isolation rather than for capacity.
Data transfer is included per compute size, in a band running from 50 GB up to 800 GB depending on tier. Laravel does not publish a per GB overage rate on the pages we read, so if your app moves serious bandwidth, ask before you commit rather than assuming it is free.
The managed database is Serverless Postgres 17 and it bills on a meter separate from your app. Cache is Valkey. Queues are metered by operation.
| Resource | Published rate | What that means monthly |
|---|---|---|
| Postgres CPU | $0.003 per hour, billed per second | $2.19 per CPU unit awake all month |
| Postgres storage, root branch | $0.15 per GB month | $3.00 for 20 GB |
| Backup history, point in time recovery | $0.15 per GB month | Doubles storage cost at equal retention |
| Managed queue operations, standard | $1 per million operations | $5.00 for 5 million |
| Managed queue operations, FIFO | Premium rate, figure not published | Ask before designing around FIFO |
| Valkey cache | By size and class, monthly caps apply | Varies, check your region |
Database compute sleeps on its own timer, which is why the tiny app in the first table pays almost nothing for Postgres. The line that surprises people is queue operations. An operation is not a job. It counts dispatch, start, completion, visibility extensions and the polling baseline, so a queue that is polled constantly accrues operations even when no jobs are running. On a Pro queue at least one worker stays up, which sets a floor under the cost of an idle queue.
Laravel does not publish a standalone rate for Reverb or for object storage buckets on the pricing pages we checked. Where a number is not published we are not going to invent one. If websockets or bucket storage are central to your app, get those rates in writing before you migrate.
It depends on awake hours, and the crossover is sharper than most comparisons admit. Below is the same four scenarios priced against a comparable server you rent yourself. DigitalOcean Basic droplet prices were verified at $6 for 1 GiB, $12 for 2 GiB, $24 for 4 GiB and $48 for 8 GiB.
| Workload | Laravel Cloud | Droplet you own | Cheaper |
|---|---|---|---|
| Awake 4 hours a month | $5.00 | $6.00 minimum, running all month | Laravel Cloud |
| Small app, always awake, 1 GB | $14.89 | $6.00 to $12.00 plus a panel fee | Close, depends on the panel |
| Production, 4 GB equivalent | $54.95 | $24.00 plus a flat panel fee | Your own server |
| Busy app plus worker, 8 GB equivalent | $118.79 | $48.00 plus a flat panel fee | Your own server |
Be honest about what the droplet column does not include. There is no managed Postgres, no CDN, no WAF, no autoscaling and no scale to zero. You are buying raw capacity and taking on the operations. A panel closes most of that gap for a flat fee by handling provisioning, certificates, firewall rules, cron, queue workers under Supervisor and atomic zero downtime releases, but it will not autoscale you through a traffic spike the way a managed platform does.
The genuinely useful conclusion is the shape of the curve, not a winner. Laravel Cloud is cheapest when your app is asleep and most expensive when it never sleeps. A fixed price server is the opposite: it costs the same at 4 hours and 730 hours. Work out which side of that curve your app sits on and the decision usually makes itself. Our own breakdown of what a Laravel VPS actually costs runs the fixed price side of this in detail.
Cost is only half a hosting decision. These are the limits that are structural rather than priced, and they are the reason some teams pay more elsewhere on purpose.
Set a spending limit on day one whatever you decide. Laravel Cloud supports a monthly cap with alerts, and compute pauses automatically when you hit it. On a usage billed platform that setting is not optional housekeeping, it is the difference between a surprise and an outage you chose.
Do this on paper before you move anything. It takes ten minutes and it is more accurate than any calculator preset, because only you know your traffic pattern.
Then run the same exercise for a server in your own cloud account with a flat fee panel on top, and compare the two totals at the awake hours you actually expect. If you want the other side of that comparison written out, we keep an honest Laravel Cloud alternative page and a Laravel Cloud versus Forge comparison that covers the control and workflow differences rather than only the money.
Ready to stop managing servers by hand? DeployManage provisions, deploys and monitors your fleet from one dashboard.
$ get started freeLaravel Cloud costs a plan fee plus metered usage. Starter is $5 a month, Growth $20 and Business $200, each including $5 of usage credit. On top of that, compute bills per second while awake, from $0.00893 an hour for Flex 512 MB. A small always awake app lands near $15 a month, and a busy production app with a worker runs closer to $119.
No, but the first month on Starter is free and every paid plan includes $5 of monthly usage credit. An app that sleeps almost all month can burn less than that credit, so the effective bill is just the $5 Starter fee. There is no permanent free tier once the first month ends.
Laravel Cloud uses usage based pricing. You pay a fixed plan fee for features, then metered rates for the resources your app consumes: compute per second while awake, Postgres CPU hours and storage, queue operations and cache. Compute stops billing when the app sleeps and wakes in under 500ms, so idle apps cost almost nothing.
Yes. Laravel publishes an interactive calculator with four scenario presets covering a hobby app, a personal blog, a SaaS MVP and a production app, and it lets you adjust awake hours, region and resource sizes. Do the arithmetic for workers and the database yourself too, because those meters are easy to leave out of an estimate.
The managed database is Serverless Postgres 17. CPU bills at $0.003 an hour per unit, charged per second, which is $2.19 a month for one unit awake continuously. Storage on the root branch is $0.15 per GB month, and point in time recovery history costs another $0.15 per GB month. Database compute sleeps on its own timer.
Postgres storage is $0.15 per GB month on the root branch, with backup history priced the same. Data transfer is included per compute size, in a band from 50 GB to 800 GB depending on tier. Laravel does not publish a standalone object storage bucket rate on its public pricing pages, so confirm that figure directly if buckets matter.
Yes. The first month on the Starter plan is free, documented as a 30 day trial. Growth, Business and Enterprise do not list a trial. Because Starter runs Flex compute that scales to zero, a trial month on a low traffic app is a realistic test of the platform rather than a throttled demo.
Flex scales to zero and comes in 512 MB, 1 GB and 2 GB sizes, so it stops billing when your app is idle and wakes in under 500ms. Pro is always on, starts at 1 vCPU with 4 GB, and unlocks autoscaling up to 10x replicas. Pro requires the Growth plan or higher.
Yes, and you should. All plans support a monthly spending limit with alerts as you approach it, and compute pauses automatically when you reach the cap. On a platform billed by usage this is the only hard protection against a traffic spike or a runaway worker turning into an invoice you did not plan for.
Only when your app sleeps. An app awake four hours a month costs the $5 plan fee, which beats any droplet running all month. Once the app is awake continuously the comparison flips: a Pro 1 vCPU instance works out near $55 a month all in, against $24 for a comparable droplet plus a flat panel fee.
Starter is $5 a month plus usage, with $5 of monthly credit and a free first month. It covers unlimited applications, environments, builds and users, Flex compute that scales to zero, spending limits and alerts, one managed queue per environment, DDoS mitigation, 10 custom domains, one day of log retention and standard support.
Forge is a flat monthly fee for the panel and you pay your cloud provider separately for the server. Laravel Cloud bundles both into one metered bill. For an always on production app, Forge plus a droplet is usually cheaper. For an app that idles most of the month, Laravel Cloud can be far cheaper.
Laravel does not publish a standalone Reverb rate on its public pricing pages, so we will not quote one. Reverb runs as part of your application compute rather than as a separately listed line on the rate card we read. If websockets are central to your product, get that pricing confirmed in writing first.
Run the same Laravel app on a server in your own cloud account, with zero downtime releases, certificates, firewall, cron, queue workers and per domain mailboxes managed from one panel at a flat fee. The box stays yours. Free to start.
# free plan · no credit card required