The free general-purpose default against the commercial enterprise standard — and an honest note on why outgrowing Postfix no longer means reaching straight for a PowerMTA licence.
In short
The story used to be simple. You started on Postfix because it is free, dependable and everywhere. Your sending grew, the per-ISP throttling got harder to manage by hand, and at some point you bought a PowerMTA licence because that was what serious volume ran on. Postfix to PowerMTA was the canonical path, and the only real question was at what scale you made the jump.
That comparison is still worth making, because the two engines genuinely sit at opposite ends: free and general-purpose at one end, commercial and built for enterprise outbound at the other. But the honest version of this page has to say something the old story left out. An open-source engine, KumoMTA, now occupies the middle of that path. Outgrowing Postfix no longer means a commercial licence is your next stop by default.
So this page does two things. It compares PowerMTA and Postfix on their own terms, so you can see exactly what the licence buys. And it tells you, plainly, where KumoMTA changes the decision — because we host all three, and steering you onto the most expensive option when a free one fits would be the opposite of useful.
Both are written in C, so the divide is not the language. It is scope, shaping and support. Postfix aims to handle every common mail job from one configuration, which makes it a fine general server and a coarse instrument for high-volume outbound specifically. PowerMTA does one thing — bulk outbound — and brings VirtualMTA, the model that binds sending streams to identities and IPs, plus per-ISP shaping behaviour the vendor has hardened against real-world quirks over a long production history.
The third piece is the one that does not show up in a feature list: support. With Postfix, you and the community are the support; with PowerMTA, a vendor is contractually on the hook with an SLA. When email is business-critical at scale, that contract is a real part of what you are buying, and for some buyers it matters as much as any shaping feature. The licence, in other words, is partly software and partly a phone number that answers.
The move rarely begins with a benchmark. It begins with a queue that backs up because every provider wants different pacing and a coarse global setting cannot satisfy them all at once. The check below — deferrals piling up with different reasons per destination — is the practical signal that outbound has outgrown a general-purpose engine.
$ postqueue -p | tail -n 1 -- 48,213 Requests deferred, growing -- $ mailq | grep -i deferred | head -3 421 4.7.0 [gmail] too many concurrent connections 421 4.7.28 [yahoo] rate limited, reduce sending speed 451 4.7.1 [microsoft] greylisted, retry later # one global pacing rule cannot answer three providers at once
None of those responses is a rejection — each is a provider asking for different treatment. Answering them individually, at volume, is the work a purpose-built engine does for you. The choice is then whether the purpose-built engine you reach for is the open-source one or the commercial one.
| Postfix | PowerMTA | |
|---|---|---|
| Licence & cost | Open source (IBM Public License), free | Commercial, per-server |
| Built for | General-purpose mail: inbound, outbound, relay | High-volume enterprise outbound |
| Language | C | C |
| Configuration | Static files (main.cf / master.cf) | VirtualMTA pools + policy files |
| Outbound traffic shaping | Coarse, per-transport | Fine, vendor-hardened over years |
| Support | Community and documentation | Vendor support with SLA |
| Volume sweet spot | Up to ~500K/day | 10M+/day |
| Choose it when | Modest volume, one general tool | Top-end volume needs vendor-backed shaping |
When leboncoin moved off Postfix as volume neared roughly 12 million a day, the destination was PowerMTA because there was little else in that class. Today the same growth meets KumoMTA first — open source, built for outbound — and many senders find it covers them without a licence. PowerMTA remains the right end of the road for the largest senders who need vendor support and its mature shaping. The diagram shows the path as it actually runs now.
You send at the top of the range — tens of millions a day — and you need a vendor contractually behind delivery with an SLA. The shaping maturity built up over years against real ISP behaviour matters most at that scale, and the standing of the established standard can carry weight in procurement and audit. If that describes you, the licence is buying something concrete.
See how it compares with the modern open-source engine directly in PowerMTA vs KumoMTA.
Your volume is modest and Postfix delivers: stay, and spend your effort on authentication and list hygiene instead. Your volume has climbed but you only need outbound done well without a licence: the open-source middle step is the move. Either way, paying for PowerMTA is not yet the answer.
The threshold itself is covered in KumoMTA vs Postfix.
The one-line rule
Outgrowing Postfix is real, but it no longer points straight at a PowerMTA invoice. Confirm the bottleneck, try KumoMTA, and pay for PowerMTA when a vendor SLA or top-end shaping is the stated need.
A vendor that only sells PowerMTA has every reason to frame outgrowing Postfix as a reason to buy. We run Postfix, KumoMTA and PowerMTA as managed infrastructure, so the advice can be the honest one: most senders leaving Postfix should look at KumoMTA before a licence, and reach for PowerMTA only when a vendor SLA or its top-end shaping is the specific requirement. Our margin does not move with that decision.
Underneath all three sits the part you cannot get from a download: dedicated IPs whose reputation we warm and manage, and EU-resident infrastructure under an Austrian entity, so the sending leg raises no cross-border transfer question in your compliance review. The engine is a choice; the dedicated, sovereign, operated foundation is the constant. The full set is laid out in the MTA comparison hub.
When outbound volume reaches the point where Postfix’s coarser shaping makes holding reputation a daily chore, and you need a vendor contractually behind the engine. Historically that move happened in the millions-per-day range. The honest update for 2026 is that you should evaluate KumoMTA first: it is an open-source engine built for high-volume outbound, so outgrowing Postfix no longer forces a commercial licence straight away. Move to PowerMTA when a vendor SLA or its top-end shaping maturity is the specific thing you need.
Only when it buys something you will use. PowerMTA’s value is vendor support with an SLA, traffic shaping refined over many years against real ISP behaviour, and its standing as the established standard in procurement and audit settings. At the very top of the volume range those are worth paying for. Below the enterprise tier, the licence buys capabilities a free engine now covers, and the better question is whether Postfix or KumoMTA fits rather than whether to pay.
The classified-ads platform leboncoin ran Postfix and migrated to PowerMTA as its daily volume approached about 12 million messages. It is a useful public marker for the shape of the high-volume path: open source carries a sender a long way, and the jump to a commercial engine happens when per-ISP traffic shaping, not raw sending capacity, becomes the bottleneck. It also predates KumoMTA, which is why the same journey today often has a middle step.
It can move a great deal of mail. The limitation is not throughput but control: its per-destination shaping is coarser than an engine purpose-built for bulk sending, so at the top of the volume range you spend more effort pacing each provider by hand to keep a clean reputation. That manual work, not a capacity ceiling, is what eventually pushes large senders toward a purpose-built engine, whether commercial or the open-source KumoMTA.
No, and that is the most important thing to know before paying. KumoMTA is an open-source engine built for high-volume outbound that sits between Postfix and PowerMTA. Outgrowing the free default no longer means a commercial licence is the only next step. The honest order is to confirm Postfix is genuinely the bottleneck, evaluate KumoMTA as the no-licence middle, and reach for PowerMTA when a named requirement makes you.
We run Postfix, KumoMTA and PowerMTA as managed infrastructure, so the recommendation can follow your volume rather than a sale. A modest sender keeps a tuned Postfix; a sender outgrowing it usually goes to KumoMTA first; a sender deep in the enterprise range who needs a vendor SLA gets PowerMTA. The dedicated IPs and EU-resident infrastructure underneath stay the same, which is why we have no reason to push the licence on you.
Tell us your volume and where the queue hurts. We will tell you whether the answer is a tuned Postfix, KumoMTA without a licence, or PowerMTA — on dedicated EU infrastructure.
