Deliverability · Economics

ESP Pricing Trends: What You’ll Actually Pay to Send Email

Email provider pricing is deliberately confusing — every provider buries costs in different tiers, charges for contacts separately from sends, and surcharges features that ought to be standard. The result is that the price you compare on a landing page is rarely the price you pay. This guide maps where ESP pricing is actually heading in 2026 and, more usefully, where the real money goes once you’re sending at scale.

How is ESP pricing structured?

Almost all ESP pricing falls into one of two models, and knowing which you’re dealing with frames every other decision. The first is volume-based pricing, where you pay either per email or for a tier that accommodates a monthly send volume — the dominant model among established providers. The second is usage-based or pay-as-you-go pricing, charging a flat rate per thousand sends with no monthly minimum, which has been gaining ground through developer-first APIs. The path of a message is the same either way; what differs is how the meter runs.

The practical distinction is what happens as you grow. Volume-based tiers tend to penalize success — every new subscriber and every extra send pushes you toward the next, more expensive bracket — whereas usage-based pricing scales linearly with what you actually send. The shape of those curves is what the chart below captures.

The guidance that holds across the market is to match the pricing model to your sending pattern rather than the reverse: usage-based APIs suit transactional sending inside an application, while flat-fee platforms suit predictable, high-volume outreach and growth.

The costs hidden beneath the sticker price

The headline plan price is only the beginning of what you’ll pay, and the gap between it and your actual bill is where most budgeting goes wrong. Dedicated IPs are the clearest example: market rates run from roughly twenty-five to sixty dollars a month per IP, and most providers won’t even let you add one until you’re sending at least fifty to three hundred thousand emails a month. The table shows representative 2026 pricing across common providers.

Beyond IPs, the add-ons stack quickly. Many providers bill contacts separately from sends, so storing a large list costs money before you send anything. List validation is often a surcharge rather than included — on the order of a dollar-plus per thousand checks — even though unvalidated lists drive the bounces that hurt deliverability. Support can be a paid tier, overage charges apply when you exceed your plan, and contact-storage overages add up. The lesson is to price the configuration you’ll actually run, not the entry plan, because the extras frequently exceed the base subscription.

The shrinking free tier

Yes, and this is one of the clearer trends of the past two years. The generous permanent free tier that defined the previous era of ESP marketing is being steadily withdrawn. One major provider converted its free plan into a sixty-day trial limited to a hundred emails a day; another cut its permanent free allowance from three thousand emails a month down to five hundred in late 2025; a third offers only a thirty-day trial with no permanent free option at all. The direction is unmistakable: evaluate-then-pay is replacing send-free-forever.

The force behind it is monetization pressure on the providers themselves, several of which now sit inside larger communications companies with revenue targets. The visible result is not only shrinking free tiers but creeping headline prices — one widely used provider raised its rates roughly ten percent across most paid tiers, with its entry plan moving up several dollars. For senders, the implication is that the cost of an ESP relationship is on an upward path, and locking in assumptions from a couple of years ago will understate what you pay today.

Why is usage-based pricing rising?

The countertrend to creeping tier prices is the rise of usage-based, developer-first APIs, and it’s a genuine shift in the market rather than a niche. These providers charge a clean rate per thousand emails with no monthly minimum — you pay when you send and nothing when you don’t — which appeals to teams whose volume is variable or who object to paying for headroom they aren’t using. The model pairs naturally with modern developer experience: clean APIs, templates treated as code, and integration that feels native to an application rather than bolted on.

This matters because it gives senders a real alternative to the penalize-growth dynamic of tiered pricing. For transactional sending in particular — password resets, receipts, notifications triggered by user actions — paying precisely per send is both cheaper and more predictable than buying volume tiers. The broader principle holds here too: the pricing model should follow your sending pattern, so a spiky, transactional workload fits pay-as-you-go while a steady promotional calendar may still suit a flat tier.

Pricing by contacts versus by sends

A pricing axis that often gets overlooked is whether a provider charges for the contacts you store or the emails you send — and the difference reshapes your bill depending on your list shape. Many providers bill contacts separately from sends, so a large but lightly mailed list incurs cost before a single message goes out. Others price purely by send volume and include unlimited contact storage, which favours senders with big lists they mail infrequently. Neither is universally cheaper; it depends entirely on the ratio of your list size to your sending frequency.

The practical effect is that two providers with similar headline send prices can produce very different bills for the same program. A team with a million stored contacts mailing them monthly pays dramatically more under contact-based pricing than send-based, while a team with a small list sending high daily volume sees the opposite. Before comparing per-email rates, work out whether your cost is driven by list size or send volume, because the right provider follows from that ratio more than from the advertised rate.

Why price and deliverability are different axes

A trend worth naming explicitly is that price and deliverability are not the same axis, and conflating them is a costly mistake. An email that lands in spam is worthless regardless of how little you paid to send it, which means the cheapest per-email rate can be the most expensive choice if its deliverability is poor. This is why providers differentiate sharply on inbox placement, and why the lowest sticker price often isn’t the best value once you account for the mail that never arrives.

The structural expression of this is stream separation. The strongest transactional providers enforce a strict split between transactional and marketing traffic, so that critical messages like password resets are never delayed or dragged down by bulk promotional sends sharing the same infrastructure. That separation is a deliverability advantage rather than a UI feature, and it increasingly shows up as a pricing and plan-design decision — you may pay for the discipline of separate streams, but it protects the mail that matters most, a theme our guide on recovering after a deliverability drop returns to.

When does ESP pricing stop making sense?

For most senders, an ESP is the right answer and the pricing is reasonable — but there’s a volume above which the model breaks down. The terminal lays out the total cost of ownership questions to ask before assuming the sticker price is your cost.

# The questions the headline price doesn’t answer
SENDS … per-email or tier? does it climb as I grow?
CONTACTS … billed separately from sends?
DEDICATED IP . required minimum volume? cost per IP?
VALIDATION … included, or a per-check surcharge?
OVERAGE … what happens when I exceed the plan?
SUPPORT … free, or a paid tier when it’s on fire at 2am?
DELIVERABILITY dashboard / inbox-placement an add-on?
# At 500K+/mo, per-email pricing often exceeds self-hosting.

The inflection point arrives around half a million emails a month and climbs from there. At that scale, per-email pricing — plus all the stacked add-ons — starts to exceed the fully loaded cost of running infrastructure yourself, and infrastructure-level options become dramatically cheaper. The cheapest infrastructure provider charges around ten cents per thousand emails, but even that looks cheap only until you add managed dedicated IPs, a deliverability dashboard that can run over a thousand dollars a month, and external warmup. The honest read is that high volume changes the economics enough to justify revisiting the build-versus-buy question, which our cost of email infrastructure guide quantifies.

What a realistic budget includes

Putting the trends together, a realistic ESP budget has four components rather than one. There’s the base plan sized to your volume; the dedicated IP or IPs once you cross the threshold that allows and warrants them; the per-feature add-ons you’ll actually use — validation, contact storage, support; and an allowance for overage, because real sending rarely matches the plan exactly. Budgeting only the base plan, as most teams do, routinely understates the true figure by a wide margin.

The strategic point underneath the numbers is that pricing trends are pushing in a consistent direction: up at the entry, with more of the cost moved into add-ons and overages, and free options narrowing. That makes it worth periodically re-pricing your setup rather than assuming last year’s plan still fits, and worth modelling what your provider will cost at next year’s volume, not this year’s. The pricing model you choose should match how you actually send — and the discipline that keeps the bill down is the same discipline that protects deliverability, since the costs you most want to avoid come from the mail that bounces or never arrives.

Does the cheapest provider win?

No — and that’s the honest conclusion. The lowest per-email rate is not the lowest cost, because deliverability, add-ons, and the penalize-growth dynamics of volume pricing all sit outside the headline number. A provider that charges more per email but lands your mail in the inbox, separates your transactional stream, and doesn’t surcharge every feature can be far cheaper in practice than a cut-rate option whose mail goes to spam. Price the outcome — mail that arrives — not the sticker.

The corollary is that the right provider depends entirely on your situation: volume, sending pattern, technical capacity, and how much a deliverability incident would cost you. Usage-based APIs suit variable transactional loads; flat tiers suit steady promotional sending; infrastructure-level providers and self-hosting suit high volume with the expertise to run them. For senders whose volume has grown past the point where per-email pricing makes sense — and who want to own their reputation rather than rent a slice of a shared pool — our PowerMTA server hosting replaces per-email economics with fixed infrastructure cost, while the deliverability discipline that no pricing plan can buy does the rest.