Search nomadLab

Transactional Email 2026: SES Changed What Happens by Default

Amazon SES is the $0.10 per thousand everyone quotes. From 21 July 2026, new accounts start on a plan that charges $0.16, and the cheap number is something you opt into. Meanwhile the break-even against a managed provider is roughly ten times higher than the usual advice says.

Updated

Every comparison of transactional email providers has the same first row: Amazon SES, $0.10 per thousand emails, cheaper than everything else by an order of magnitude. That number is still on the pricing page. It is no longer what a new account gets.

SES now sells three plans. Essentials charges $0.16 per 1,000 emails for the first 10 million, Pro charges $0.22 plus $105 per account per region per month, Enterprise charges $0.23 plus $500. The old pay-only-for-what-you-use rate is now labeled “à la carte,” and the footnote under the plan table says that new SES accounts, and account-and-region combinations with no metered activity since 1 June 2025, “will start on the Essentials plan beginning July 21, 2026.” You can switch to à la carte at any time. You just have to know to do it.

Sixty percent is not a large absolute difference at small volume. It is a large difference in what the comparison table means, because the number everyone quotes is now the number you get by acting rather than the number you get by default.

What the five actually cost

A bar chart of monthly cost at 100,000 emails per month. Amazon SES à la carte is $10. Amazon SES on the Essentials plan is $16. Resend Pro is $35. SendGrid Pro 100K is $89.95. Mailgun Scale is $90. Postmark Basic is $177. 100,000 emails a month, per provider Amazon SES, à la carte $10 Amazon SES, Essentials $16 Resend Pro $35 SendGrid Pro 100K $89.95 Mailgun Scale $90 Postmark Basic $177
Off vendor pricing pages on 23 August 2026. Postmark's figure is its $15 base for 10,000 plus 90,000 at $1.80 per thousand; its Platform plan works out to $126 at the same volume.

At lower volumes the shape is different. Below is the same set at three volumes, using each vendor’s cheapest plan that covers it.

Provider1,000 / mo10,000 / mo100,000 / mo
Amazon SES (à la carte)$0.10$1$10
Amazon SES (Essentials)$0.16$1.60$16
Resendfree$20$35
Mailgunfree$15$90
Postmark$15$15$177
SendGrid$19.95$19.95$89.95

Resend’s free tier is 3,000 emails a month with a 100-per-day cap, and the daily cap is what pushes people to upgrade rather than the monthly one. Mailgun’s free tier is also 100 a day and is a permanent plan rather than a trial. Postmark’s free tier is 100 emails a month, which is an integration test, not a plan.

SendGrid has no free tier at all any more, and this is worth dating correctly because the internet keeps calling it recent news. Twilio retired the SendGrid free plan on 27 May 2025, gave existing free users 60 days, and then paused their sending. Paid Email API plans start at $19.95 a month for Essentials and $89.95 for Pro.

Postmark’s cheapest plan stops being cheapest at 13,000 emails

Postmark restructured into a base price plus a per-thousand overage, and the three tiers trade base against rate: Basic is $15 with overage at $1.80 per thousand, Pro is $16.50 at $1.30, Platform is $18 at $1.20. All three include the same first 10,000 emails.

That means the tiers cross. Basic and Pro are equal at 13,000 emails a month, and above that the $16.50 plan is the cheaper one. Pro and Platform cross at 25,000. So the cheapest-looking plan is the wrong choice for anyone sending more than about a thousand emails a day, and the pricing page does not tell you that. Run the arithmetic on your own volume: base plus 1.8 times thousands-over-ten versus base plus 1.3 times the same, and take whichever is lower.

The SES break-even is about ten times higher than people say

The usual rule of thumb is that SES starts to pay for itself somewhere around 200,000 emails a month. Check that against the table.

At 100,000 emails, the most expensive managed option in this comparison is Postmark Basic at $177, against $10 for SES à la carte. The saving is $167 a month. That is under two hours of a loaded engineer’s time in most places, and SES setup is a day or two before you count the maintenance: DKIM, SPF, and a custom MAIL FROM you configure yourself, bounce and complaint handling through SNS that you parse and act on, your own suppression list, and getting out of the sandbox before you can email anyone who has not verified their address.

Against Resend, the option most small teams are actually choosing, the saving at 100,000 emails is $25 a month. There is no version of the arithmetic where that pays back a day of setup.

The volume where it does start to matter is higher. Resend’s Scale plan runs $90 at 100,000 emails and $1,150 at 2.5 million. SES à la carte at 2.5 million is $250, so the gap is about $900 a month, which is roughly a day of engineering time every month. That is where SES stops being a hobby and starts being a decision. On the Essentials plan the same volume is $400, so the gap narrows to $750 and the break-even moves up again.

None of that makes SES a bad choice. It makes SES a choice you take because you are already in AWS and comfortable with SNS and CloudWatch, not one you take because a table said ten dollars.

Dedicated IPs, and what they cost everywhere

If you send enough that shared-pool reputation is a real risk, the dedicated IP line is where the bills diverge again.

SES charges $24.95 per month per standard dedicated IP, or $15 per month per account plus $0.08 per thousand emails for managed dedicated IPs, which warm and scale themselves. Bring-your-own-IP is also $24.95 per IP, with a 256-address minimum, so a floor of $6,387.20 a month. Resend charges $30 a month for a dedicated IP and only offers it above 3,000 emails a day on the Scale plan, which is a sensible gate. SendGrid and Mailgun include or sell them on their higher tiers.

A dedicated IP you neglect is worse than a well-run shared pool, which is the argument for the managed variants. If nobody on the team is going to watch the warm-up curve, do not buy the raw one.

Deliverability is the part the table cannot show

On entry plans you send from IPs shared with hundreds of other customers, and their behavior becomes partly your reputation. That is the mechanism behind most “our password resets started going to spam” incidents, and there is no pricing-page column for it.

Postmark’s answer is to separate transactional and broadcast streams onto different infrastructure and to remove senders who damage the pool, which is what the premium buys. Resend has a good reputation and gates dedicated IPs behind real volume. SES has strong underlying IP reputation and hands you all of the reputation work.

The rule that applies whatever you pick: keep transactional and marketing on different sending domains, ideally different providers, and set the second one up before you need it. It is a few DNS records now. Untangling it after a campaign has damaged the domain your password resets go out on is a much worse week.

What I would do

Small project on a JavaScript or TypeScript stack: Resend. The free tier covers you until the 100-a-day cap bites, Pro at $20 covers 50,000, and React Email is the nicest template story in the group.

Email that has to arrive: Postmark, on the right tier. Pick by the crossover arithmetic above, not by which base price looks smallest.

Already in AWS, sending in the millions: SES, and check which plan your account is on before you budget anything. If it was created recently, or has been idle in that region since June 2025, the number in your spreadsheet is probably $0.10 and your bill is going to say $0.16.

Migrating off SendGrid because the free plan died: the move is usually an afternoon of DNS records and one SDK swap. The same reasoning applies here as in the object storage layer, where the cheap bring-your-own option only wins once you already have the in-house competence to run it.

Prices came off the AWS SES pricing page, the four vendor pricing pages, and Twilio’s changelog on 23 August 2026. The one to re-check before you commit is the SES plan footnote, because 21 July 2026 has already passed and which plan your account defaulted to is not something the comparison tables have caught up with.

Keep reading