Laravel Cloud Gets Scale-to-Zero: Your Idle Apps Stop Burning Money

If you have been running staging environments, hobby projects, or low-traffic production apps on Laravel Cloud, you have probably noticed one annoying thing: even when nobody is using the app, the compute keeps running and the bill keeps climbing. That changes with the latest update.

Laravel Cloud just shipped Scale-to-Zero Flex compute, and it is exactly what it sounds like. Your entire application stack — web, workers, the works — can now sleep when there is no traffic and wake on the first incoming request.

Scale-to-Zero: The Details

When your app is idle, Laravel Cloud now spins the stack down completely. No compute running, no charges accumulating. When a request comes in, the stack wakes and handles it.

The key metric here is cold start performance: under 500ms. Cold starts were historically the pain point of scale-to-zero hosting. Platforms that offered similar features often had cold starts in the 2-5 second range, which makes the first request feel broken rather than just slow. The Laravel team has brought this down to sub-500ms, which is fast enough that most users will not notice the difference.

The announcement puts the improvement at 20x faster cold starts compared to previous approaches. The team ran a developer survey earlier this year, followed up with 20 individual interviews, and the consistent feedback was that unpredictable billing and wasted spend on idle apps were the two biggest friction points. Scale-to-zero directly addresses both.

New Four-Tier Pricing

Alongside scale-to-zero, the pricing structure has been reorganized into four tiers:

Every paid plan includes $5 in monthly usage credits drawn from a single shared pool. That pool covers everything: compute, database, cache, storage, bandwidth, queue operations, and custom domains. You are not managing separate budgets per service category, which was another common complaint from the previous pricing model.

Spending Limits

The new update also introduces monthly spending limits at the org level. You set a cap, and Laravel Cloud will not bill you beyond it. This addresses the specific anxiety that comes with usage-based pricing: the fear of leaving something running and coming back to an unexpected bill.

For teams managing multiple apps, this is meaningful. You can set a global limit for your organization and not worry about a runaway job or a misconfigured cron eating through your budget before anyone notices.

What This Means for Staging and Side Projects

The combination of scale-to-zero and the $5/month Starter plan makes Laravel Cloud significantly more compelling for:

Staging environments that sit idle 95% of the time. Previously, running a staging environment cost nearly as much as production. With scale-to-zero, a staging app that gets hit a few times a day barely registers on the bill.

Side projects and demos that have real traffic patterns — busy some days, nothing for a week. You pay for what you use rather than for the uptime you reserved.

Laravel Horizon and queue workers are worth noting specifically: for long-running queued jobs, the docs recommend using Managed Queues rather than a queue worker that scales to zero, since a job running when the stack sleeps would be interrupted. For short, fast jobs, scale-to-zero workers are fine.

A Note on Cold Starts for Production

For high-traffic production apps, scale-to-zero is probably not what you want. If you are serving thousands of requests per minute, the stack never goes idle anyway, and keeping compute warm is better. Scale-to-zero is most valuable for apps with predictable traffic gaps.

The Laravel Cloud team has made it easy to configure which environments use Flex compute and which stay on always-on instances, so you can make the right call per environment.

Full details on the scale-to-zero update are on the Laravel blog. The new pricing is already live.

Sources: Laravel Blog · Laravel News · Laravel Cloud