Do Frappe apps scale?

This is one of the first questions serious buyers ask, and it deserves an honest answer rather than a reflexive "yes."

The honest answer is: it depends on your specific workload - and Frappe is more scalable by default than most people expect. A confident "yes, it scales infinitely" from any vendor should make you skeptical, because real scalability is never unconditional. What matters is understanding your constraints, knowing where the limits actually are, and having a clear plan for the rare cases that need engineering. This section gives you that.

What "scale" actually means

When people say "does it scale," they're usually conflating two different things, and separating them makes the conversation precise:

• Latency - how fast the system responds to a single action. Opening a document, saving an invoice, running a report. When users complain about performance, they're almost always complaining about latency ("saving takes ten seconds").
• Throughput - how much the system can process over time. Ten thousand invoices a day, payroll for tens of thousands of employees. When users worry about scalability in advance, they're usually thinking about throughput.

The two are related - very high latency on an operation drags down throughput - but they're diagnosed and solved differently. A good scalability conversation asks about both: not just "how many users," but "how many doing what, how often, and how fast does each action need to be?"

More scalable than expected - some real examples

Without naming customers, here is the kind of scale Frappe applications run in production today:

• A payroll run for 30,000+ employees.
• A stock-broking platform processing millions of transactions per day.
• A deployment spanning 500+ warehouses across 13 companies and 40 branches, sustaining continuous transaction volume.
• A migration that processed 100,000+ invoices in a matter of hours, replacing a legacy enterprise system.

These aren't theoretical benchmarks; they're live systems. They don't prove that any workload will work unchanged - they prove that Frappe's architecture reaches serious enterprise scale, and that when a specific workload pushes a limit, there's a known path through it.

How to read the rest of this section

The reality of scaling is that it's a layered playbook, moving from easy to expert:

1. For most workloads, it simply works at the default configuration.
2. When you need more, you scale up resources - often in minutes.
3. When resources stop being the answer, you find and fix the bottleneck - usually in the database.
4. For unusual or extreme cases, there are architectural options and expert help.

The pages that follow walk through this:

How scaling works architecturally

The scaling playbook

How scaling relates to cost

Honest limits and when to get help

If you're a smaller business: most of this section describes what's possible at the upper end, not what you'll need to do. At your scale, Frappe runs comfortably on default configurations - you can read this for reassurance that you won't outgrow the platform, and skip the technical depth unless you're curious.