At 100K Emails a Month,the Send Rate Is the Small Number

The short answer

Every figure in this post is either taken from a published pricing page or computed from published rates, and the arithmetic is shown so you can check it. AWS rates were verified against the SES pricing page in August 2026. Competitor pricing was verified against live pricing pages on July 13, 2026.

What 100,000 emails actually costs

Amazon SES has four billing modes since July 2026. Here is each one priced at exactly 100,000 emails a month. Monthly fees are charged per account, per Region.

One conclusion falls straight out of that table and holds at every volume: à la carte is always the cheapest way to buy raw sending. Its $0.10 rate is lower than the best marginal band any subscription plan reaches — Essentials bottoms out at $0.11 above 100M, Pro at $0.12 — and it carries no monthly fee. There is no crossover point. If all you want is delivery, no SES plan ever beats à la carte on price.

That does not make the plans a trap. They bundle real products with real standalone prices — Essentials includes Virtual Deliverability Manager, which AWS otherwise sells at $1,250 a month for the global tier. At 100K a month, Essentials costs you $6 more than à la carte. If you use VDM, that is the cheapest VDM will ever be. If you don't, you are paying a 60% premium on sending for a dashboard you never open.

The rest of the bill, in full

"SES is $10" is true and incomplete. SES delivers email; it does not store your event history, and every open, click, bounce, complaint and delivery has to land somewhere. Here is the whole monthly bill for 100,000 emails through a standard SES event pipeline — EventBridge to SQS to Lambda to DynamoDB — in us-east-1, at 8 event types per email and 90-day retention.

$2.50 of infrastructure on top of $10 of sending. That is the real floor for a 100K-a-month transactional sender with full event history in their own account. Hold that number in mind, because every optimization below gets measured against it.

Step 1: Find out which plan you're on

On July 21, 2026 AWS started assigning the Essentials plan to every new SES account, and to every account-and-Region pair with no metered SES activity since June 1, 2025. Accounts that were actively sending were not repriced. The unit is not "account" — it is account × Region.

So the answer to "which plan am I on" is not one answer. It is one answer per Region, and the Regions most likely to be sitting on Essentials are exactly the ones nobody audits: the staging account, the second Region you expanded into last quarter, the disaster-recovery Region you provisioned and never used. At 100K a month the difference is $6. At 1M it is $60. It compounds silently and it never shows up on your bill as anything other than a slightly larger number.

# Read the plan for every Region you send from, and model
# your real volume against all four billing modes
wraps email plan

# Model a specific volume
wraps email plan --volume 100000

# Move a Region back to à la carte
wraps email plan --region us-east-1 --set alacarte

If you'd rather do it by hand, the plan is readable through the SES GetAccount API and writable through PutAccountPricingAttributes, once per Region. One detail is worth knowing before you plan a migration window: an account that AWS defaulted onto Essentials can cancel back to à la carte with immediate effect. Every other downgrade waits for the next billing cycle.

Step 2: Find what costs more than your sending

At $10 of sending, the bar for "this line item dwarfs my sending" is very low. Four things routinely clear it.

Step 3: Price the switch honestly

Here is 100,000 emails a month across the providers most transactional senders are actually choosing between, computed from published plan prices and overage rates as verified on July 13, 2026.

Two things are worth pulling out of that table. First, the spread between the cheapest and most expensive way to send the same 100,000 emails is 17×, and nothing about the messages changed.

Second, look at the three Postmark rows. At exactly this volume their most expensivebase plan is the cheapest total: Platform at $18 a month with a $1.20 overage comes to $126, while Basic at $15 a month with a $1.80 overage comes to $177. Buying the higher plan saves $51 a month. That inversion is a general property of include-plus-overage pricing, not a Postmark quirk — if your provider prices that way, recompute your own tier at your own volume before you assume you're on the right one.

But at 100K a month, moving to SES saves $25 to $167. That is real money and it is not, on its own, a good reason to migrate email infrastructure. The actual reason is the shape of the curve:

À la carte is a straight line at $0.10 per 1,000 forever. Managed providers price in tiers with step changes, and the steps get steeper. If 100K a month is where you are today and 1M is where you'll be in eighteen months, migrate for the 1M number, not the 100K one — and migrate while the volume is small enough that warming and cutover are easy.

The standard advice, priced out

Search this question and you'll get the same four suggestions everywhere. Three of them are good advice for reasons that have nothing to do with cost, and it's worth being precise about which is which.

"Reduce the number of emails you send"

This is routinely called the highest-ROI optimization. The value of deleting an email is exactly your marginal per-email rate, so here is what cutting 30,000 emails a month — a 30% reduction, which is an ambitious engineering project — is worth in each place you might be sending from:

An 18× spread on identical work. On SES à la carte, deleting nearly a third of your email volume saves $3.00 a month — you cannot fund a sprint with that. On Postmark's Basic overage it saves $54.00 a month, and the same sprint pays for itself. Multiply before you scope.And if you find duplicate sends, race conditions firing two password resets, or a retry loop with no ceiling, fix them anyway — those are bugs your users can see, which is a much better reason than $3.

"Suppress bounces and complaints immediately"

Do this, but not to save money. A 3% hard bounce rate on 100,000 sends is 3,000 wasted emails, worth $0.30 a month. The reason to run a tight suppression list is that bounce and complaint rates are what get your sending paused, and a paused account costs you every email, not 3% of them. AWS maintains an account-level suppression list at no charge and it is on by default — the work is making sure your own retry logic respects it rather than reintroducing addresses it already dropped.

"Separate transactional from marketing"

Correct, and also not a cost measure — it's a reputation measure. Poor engagement on marketing sends drags the reputation that your password resets depend on. On SES the mechanism is configuration sets and, if volume justifies it, separate subdomains. The cost impact is indirect but large: it is the difference between your receipts landing in the inbox and landing in spam.

"Don't self-host SMTP"

Right, and the arithmetic is blunter than the usual framing. At 100K a month there is no sending cost left to recover — the entire prize is $10. What you would take on to win it is IP warming, bounce and complaint processing, DKIM, SPF and DMARC operations, queueing, retry policy, blacklist monitoring and deliverability triage. That is not a $10-a-month problem at any company.

The order to do this in

What this costs with Wraps in it

It would be a strange post to argue that hidden line items are the problem and then hide ours. Wraps deploys the SES pipeline priced above into your AWS account — the $12.50 lands on your bill, from AWS, with no markup from us. What we charge for is the platform on top: the dashboard, templates, workflows, broadcasts and segments.

The Free tier includes the CLI and the TypeScript SDK with 5,000 tracked events a month, so a 100K-a-month sender who wants infrastructure and nothing else pays us nothing. If you want the full send history in the dashboard, 100,000 emails generates enough tracked events to need Starter at $19 or Growth at $79, depending on how many event types you record.

Which means the honest full-stack number at 100K a month is $12.50 to AWS plus $0–79 to us — and worth comparing against the $35–177 in the provider table, with the difference that the AWS half of it keeps costing $0.10 per 1,000 at 10M.