Resend Pricing Guide for 2026

Who this guide is for

This guide is for the engineering-led team that already thinks of email as a system rather than a campaign: developers wiring receipts and notifications through an API, technical founders who want the bill to track usage the way their other infrastructure does, and product teams whose newsletters and workflows sit alongside transactional traffic rather than far away from it. It assumes you are comfortable budgeting three meters instead of one, because that comfort is the actual entry requirement for Resend's model.

It also serves the buyer who has to defend the choice to someone else. If you need to show a founder, a finance lead, or a client why the invoice moved this month, the sections below give you the arithmetic: what each meter measures, where each one bites, and which structural difference matters most as the list grows. Every figure comes from the official Resend pricing page as verified on September 17, 2026.

If you are not yet sure whether three meters or one is right for you, read the pricing geometry section before the tier walkthroughs. Most bad fits are not discovered at checkout; they are discovered six months later when a team realizes it was forecasting one axis while paying for three.

Read this guide closely if

• Engineers own the sending stack and prefer usage-shaped pricing to audience-shaped pricing.
• Transactional volume is predictable but marketing contacts and automation runs are not.
• You want a developer-grade API and are willing to accept a contact meter for the marketing side.
• You are forecasting a launch spike and need to know exactly what an overage costs per thousand.
• You would rather model three small meters precisely than guess at one large bundled number.

Look elsewhere first if

• One bundled invoice with no usage arithmetic is a hard requirement for your finance process.
• Billing-triggered dunning and expansion flows are the core of the program rather than a later addition.
• Your stored contacts vastly exceed your monthly engaged senders, which makes any contact meter expensive by construction.
• Journey complexity across many channels is the primary reason you are buying at all.

None of those exclusions imply a weaker product. They describe a shape of team for which a different geometry — pooled volume, bundled journeys, or contact-unlimited sending — removes whole categories of forecasting work instead of adding meters to it.

Three meters, not one price

The single most important thing to understand about Resend pricing is that no single subscription covers everything. Transactional sending, marketing contacts, and automation runs meter independently, so a team running receipts plus newsletters plus workflows pays across up to three axes. That separation suits developers because each axis maps to a distinct system concern: application mail volume, audience size, and workflow executions. It punishes buyers who budget from one headline number and discover the other meters only on the invoice.

Start by inventorying your program along those three axes before comparing vendors. Count monthly transactional sends including receipts, invites, resets, and notifications, since these scale with product usage rather than marketing ambition. Count marketable contacts separately, since dormant profiles still occupy marketing tiers. Then estimate automation runs, the sneakiest meter, by multiplying enrolled contacts by workflow steps and re-entries. Only with all three numbers can Resend's structure be compared fairly against pooled or profile-metered platforms like Customer.io.

Pricing geometry explained

Resend's geometry is three independent meters stacked on one account. The transactional meter counts emails sent from the application; the marketing meter counts stored contacts; the automation meter counts workflow runs. None of the three is aware of the others, so an account can sit comfortably inside two allowances while the third quietly crosses a boundary — which is precisely why forecasting from a single headline number fails on this platform.

Each meter also grows for its own reasons. Transactional volume scales with product usage: more signups means more verification mail, more activity means more notifications, and a successful launch moves that axis without anyone choosing to. Marketing contacts scale with acquisition and list hygiene: every import and every dormant trial adds to the meter whether or not that record will ever be mailed again. Automation runs scale with ambition: each new workflow, each additional step, and each re-entry rule multiplies consumption on top of an audience that was already counted.

The growth story differs sharply from pooled or contact-unlimited models. On a send-only meter, a growing list is inert until it is mailed; on Resend's marketing meter, a growing list is a growing bill the moment it is stored. That is not a defect — it is the price of treating the audience as a managed asset with its own meter — but teams should enter it knowingly, because the contact axis starts taxing before the send axes do.

The favorable reading is that all three meters are legible and independently controllable. Engineers can reduce transactional waste with better targeting, marketers can suppress cold records to hold the contact meter, and operators can audit workflow steps to trim runaway runs. On a bundled plan, none of those levers map to a line item; on Resend, each one does, which rewards disciplined teams and penalizes absent-minded ones.

What a growing list does to the bill

Model the three axes separately, at signing, at twelve months, and under a stress case where the free-user database balloons. A program can triple its stored contacts while keeping sends flat, and the marketing meter will triple anyway; another program can hold contacts at a thousand while transactional traffic grows tenfold, and only the first meter reacts. The right question is never "what does Resend cost" but "which of my three growth curves hits a boundary first."

The automation meter deserves extra pessimism in forecasting because it compounds silently. Enrolled contacts multiplied by workflow steps is a lower bound, not an estimate, once re-entry rules, date triggers, and branch rejoins exist. Treat every new playbook as a recurring monthly commitment on that meter and require the projected run count at approval time — the same discipline engineers apply to API rate budgets.

The table's most important row is the dormant database one: it is the only shape where nothing is being sent, nothing is running, and the bill still rises. That is the structural cost of contact-based marketing pricing, and it is the single line to stress-test against your own list before committing.

Read the rest as arithmetic you can redo yourself. Transactional overage at $0.90 per thousand and automation overage at $0.0015 per run are both linear and predictable; the contact meter is the only axis whose growth you do not fully control, because acquisitions and imports arrive from outside the email team.

Transactional pricing, verified

The transactional free tier allows 3,000 emails per month with a daily cap of 100, which serves prototypes, side projects, and staging environments genuinely well. A pre-launch SaaS sending welcome mail, verification codes, and founder-led onboarding can live inside this tier for months. The daily cap is the real constraint: a launch-day spike of signups or a bulk password reset can hit 100 sends before lunch, so teams approaching launch should plan the Pro step rather than discovering the ceiling mid-spike.

Transactional Pro at $20 for 50,000 emails per month covers serious early-traction volume, roughly 1,600 transactional sends daily, with overages at $0.90 per additional 1,000 emails if you keep sending past the quota. That overage rate keeps a one-month spike predictable: another 10,000 transactional emails adds about nine dollars, so spikes translate to single-digit surprises rather than invoice shocks. A sustained climb is a different calculation, because Resend's pricing slider, read on October 7, 2026, steps Pro up to $35 at 100,000 emails, and Scale starts at $90 for 100,000 and reads $160 at 200,000. A team that stays above 50,000 every month should price its own volume on the slider rather than stacking overage on the $20 base. Teams should still separate transactional from marketing volume mentally, because marketing broadcasts draw on the contact-based meter instead, and conflating the two is the most common forecasting error we see.

For agent-run operations, transactional mail is usually the safest workload to automate first. Receipts and notifications follow deterministic templates with low reputational risk, and agents calling a clean send API with webhook delivery confirmation fit naturally into approval-light flows. Our discussion scenarios on approval thresholds and audit trails map this progression in detail.

Marketing pricing, verified

Marketing pricing follows contact-based logic: a free tier covering 1,000 contacts, then Pro at $40 for 5,000 contacts. Small newsletters and early lifecycle programs fit comfortably, and the free tier genuinely supports validation before spending. The structural consideration is that contacts include the unengaged: imported legacy lists, dormant trials, and cold subscribers all occupy the meter whether or not they ever open anything.

Compare this honestly against contact-unlimited alternatives. A SaaS business with 20,000 stored profiles but only 4,000 engaged senders pays for storage on Resend's marketing meter, while a send-metered model charges only for activity. Conversely, a disciplined sender with a small engaged list and heavy transactional volume may find Resend's split meters cheaper than any pooled plan, since transactional overages at $0.90 per thousand are gentle. Neither structure is universally cheaper; the winner depends on your ratio of stored contacts to monthly sends.

List hygiene therefore carries direct dollar value on Resend marketing tiers. Quarterly suppression of hard bounces, chronic non-openers, and role addresses keeps the contact meter aligned with reachable humans. Agents can run these audits well under supervision, and several of our discussion scenarios cover permission design for exactly this maintenance work.

Automations meter: runs add up

Automations include 10,000 runs before metering at $0.0015 per additional run, and this meter surprises teams most often. A run fires per enrollment traversal rather than per campaign, so a three-step welcome flow enrolling 4,000 new users monthly already consumes 12,000 runs before any newsletter automation exists. Re-entry loops, date-based rejoins, and multi-branch dunning multiply runs further.

The math stays friendly at moderate scale: 10,000 extra runs cost about fifteen dollars, so even a 50,000-run program adds roughly sixty dollars beyond the allowance. But agent-designed workflows that enroll broad segments into multi-step flows can cross the allowance quickly, and each new playbook should ship with a run estimate the way engineers ship with a latency budget. Require agents to report projected runs at approval time, and review automation consumption monthly alongside sends and contacts.

Hidden fees and fine print

Resend's surprises are rarely hidden in legal language; they are structural, sitting in the gap between what a team thinks it bought and which meter actually moved. The list below covers the four patterns that produce the first confusing invoice, plus the edges that were outside our verified source set and therefore have to be checked live.

The split between transactional and marketing

The most common forecasting error is treating transactional and marketing sending as one pool. They are separate meters with separate allowances: application mail draws on the transactional side, audience mail draws on the contact side, and a month where both spike at once produces two charges rather than one combined overage. Teams migrating from a single-pool vendor often discover this in their first launch month, when a marketing broadcast and a transactional spike land together.

The discipline that prevents it is bookkeeping, not negotiation. Tag every send at the API level by purpose, keep a rolling monthly total for each meter, and set a soft threshold at eighty percent of each allowance so the overage conversation happens before the invoice. Our discussion scenarios on cost attribution cover exactly this tagging pattern.

Automation runs are a third bill, not a feature

Automations include 10,000 runs and then meter at $0.0015 per run, and the framing matters: runs are consumed by workflow traversals, not by campaigns sent. A three-step flow enrolling 4,000 people per month consumes 12,000 runs on its own, which is already past the allowance before any other workflow exists. Re-entry rules and multi-branch logic multiply this further, and none of it is visible on a page that shows contacts and sends.

Treat run consumption as a budget line reviewed monthly, with each new playbook carrying a projected run count at approval. At the verified rate, forty thousand runs beyond the allowance cost about sixty dollars — affordable when deliberate, alarming when it arrives unexamined alongside a contact-tier step.

Daily caps and contact floors

The free transactional tier's 100-per-day cap is a launch-day hazard rather than a monthly constraint: 3,000 emails per month is generous, but one signup surge can exhaust the daily allowance before the afternoon. On the marketing side, the binding constraint is storage rather than activity — 1,000 free contacts and $40 for 5,000 on Pro mean that imports, dormant trials, and cold segments occupy paid capacity whether or not they are mailed.

Both patterns reward proactive steps: move to transactional Pro before a known launch window rather than during it, and suppress or sunset cold records quarterly so the contact meter tracks reachable humans instead of an archaeological dig.

Tier edges and packaging

Our verified figures stop at transactional Pro at $20 for 50,000 with $0.90 per thousand overage, marketing Pro at $40 for 5,000 contacts, and automation runs at $0.0015 after the first 10,000. Beyond those sit the transactional slider readings from October 7, 2026: Pro $35 at 100,000 emails, Scale $90 at 100,000 and $160 at 200,000. Marketing tiers beyond 5,000 contacts, transactional volumes between and beyond those slider points, annual discounts, seat pricing, and any add-ons were not rendered in our source set, so treat each as a live-verification item rather than an assumption.

Ask three questions in writing before signing: how overages are billed and when, whether a contact-tier step is retroactive to the current cycle, and what happens to scheduled automation runs if the account changes tier mid-month. These are ordinary clauses with extraordinary invoice potential.

A pre-signing checklist for the three meters

• Tag every send as transactional or marketing and track both meters on a rolling monthly dashboard.
• Estimate automation runs as enrollees multiplied by steps, then add re-entry and branch expansion.
• Time plan steps for known launch windows instead of discovering the 100-per-day free cap mid-spike.
• Apply suppression quarterly so the contact meter measures reachable humans, not dormant records.
• Confirm current marketing tiers above 5,000 contacts, and read the slider price for your own transactional volume above 50,000 on the vendor's page.
• Re-verify every figure at Resend pricing on decision day; all figures here were verified September 17, 2026.

Worked scenarios using only verified numbers

Scenario A: developer SaaS, transactional-heavy

A developer tool sending 30,000 transactional emails monthly plus a 800-contact changelog newsletter fits transactional Pro at $20 for 50,000 with headroom, marketing free at 1,000 contacts, and likely stays inside 10,000 automation runs. Total metered spend is roughly twenty dollars plus any automation overflow, an excellent outcome that explains Resend's developer loyalty. The equivalent pooled plan elsewhere would need to beat twenty dollars all-in to compete.

Scenario B: growing newsletter plus workflows

A team with 4,500 marketing contacts, 20,000 transactional sends, and 25,000 automation runs monthly lands on marketing Pro at $40 for 5,000 contacts, transactional Pro at $20 for 50,000, and 15,000 automation overage runs at about twenty-two dollars and fifty cents. Combined metered spend is roughly eighty-two dollars monthly before any higher marketing tiers. At this shape, contact-unlimited pooled alternatives deserve a serious look, since contact growth alone will force the next marketing step while automation appetite keeps compounding.

Scenario C: launch spike month

A launch driving 60,000 transactional emails in one month exceeds the 50,000 Pro allowance by 10,000, adding about nine dollars in pay-as-you-go overage if the account simply stays on Pro, while the 100-per-day free cap is long gone. That arithmetic holds for a one-month spike; if 60,000 were the new normal, the number to compare is the price the slider shows at that volume, because the slider already reads only $35 at 100,000. The lesson is structural: keep Pro active through launch windows rather than gambling on free-tier daily caps during the highest-visibility week of the year. Nine dollars of overage beats a throttled signup flow by orders of magnitude.

Cost scenarios: three monthly invoices

The worked scenarios above map shapes to tiers; these three build the actual statement. Each uses only verified figures from this guide and the pricing index: transactional free of 3,000 per month at 100 per day, transactional Pro at $20 for 50,000 with $0.90 per 1,000 overage, marketing free at 1,000 contacts, marketing Pro at $40 for 5,000 contacts, and automations including 10,000 runs before $0.0015 each.

Month three: engineers-only stack

The product sends 48,000 transactional emails — verification, receipts, notifications — to a small early audience, holds 800 marketing contacts for a changelog list, and runs a single welcome flow consuming 5,000 automation runs. Transactional volume sits under the 50,000 Pro allowance, so the charge is $20; the contact count is under the 1,000 free tier, so marketing adds nothing; runs sit inside the 10,000 included, so automation adds nothing either.

The month's honest total is about twenty dollars, and the structural observation is that this shape barely touches two of the three meters. Teams at this stage should still instrument all three anyway, because the meter that will bite first is usually the one nobody is watching — here, contacts, since every product signup is also a potential changelog subscriber.

Month seven: all three meters active

Transactional sending settles at 42,000 emails, the marketing list grows to 4,800 contacts, and three workflows now consume 33,000 automation runs. The transactional charge remains $20 under Pro, marketing steps to Pro at $40 for 5,000 contacts, and runs exceed the allowance by 23,000, which at $0.0015 each adds about thirty-four dollars and fifty cents.

Combined monthly spend lands near ninety-four dollars and fifty cents, and roughly a third of it comes from a meter — automation runs — that appears nowhere in the headline pricing. This is the invoice shape that surprises teams, and it is entirely predictable once run estimates are attached to every new playbook.

Month ten: overage month after a launch

A launch pushes transactional volume to 60,000, exceeding the 50,000 Pro allowance by 10,000 and adding nine dollars at the $0.90-per-thousand overage for a month that stays on Pro. Marketing stays lean at 4,200 contacts for forty dollars, while launch-period workflows and re-entries drive automation runs to 68,000, of which 58,000 bill beyond the allowance at about eighty-seven dollars.

The resulting month costs roughly a hundred and fifty-six dollars: twenty for Pro, nine of transactional overage, forty for contacts, and eighty-seven of automation overage. Nothing in that total is punitive — every rate is linear — but about ninety percent of the overage comes from automation runs, which is why this guide insists on run budgets at approval time rather than invoice time.

Set those totals beside reference points such as Brevo's free 300 emails per day with Starter at $9 and Standard at $18, and the comparison is not about which sticker is lower. It is about which meter matches the way your program grows: three usage axes here, profile pricing elsewhere, and volume steps in infrastructure tools.

The decision rule that survives all three framings is the same: model your own numbers, pick the geometry whose failure mode you can live with, and re-verify the live page before the first invoice rather than after it.

One further note for teams comparing across vendors: the totals above exclude anything our source set did not render — annual discounts, seats, add-ons, and marketing tiers beyond 5,000 contacts — so treat them as complete only for the lines shown, and confirm the rest on the vendor page before these figures enter a budget document.

How Resend compares with lifecycle platforms

Resend wins developer experience and transactional API purity. Teams that also need lifecycle depth should budget a second system or their own orchestration layer, then price the integration and suppression sync alongside Resend's meters.

Against Customer.io, Resend is simpler and cheaper at low complexity while Customer.io earns its keep on multi-channel journey expressiveness; our Customer.io pricing guide covers that metered structure. Against Brevo and Klaviyo reference points, Brevo's free 300-per-day tier and $9 to $18 entry steps undercut on raw budget, while Klaviyo's free 250 profiles plus 500 emails monthly targets commerce specifically.

How Resend fits SaaS teams

For a SaaS business specifically, the comparison is less about email quality and more about which events drive the program. Resend's meters are organized around sending mechanics: application volume, audience size, and workflow executions. Billing events and trial sequences require explicit integration design around those primitives.

The economic difference shows up first on the contact axis. A SaaS with 30,000 stored users, most of them dormant, occupies real capacity on Resend's marketing meter while costing nothing to store on a contact-unlimited model; the same product sending 40,000 transactional emails pays comfortably under Resend's transactional Pro at $20 for 50,000 either way. Whether the second axis matters depends entirely on your ratio of stored users to engaged subscribers — which is why this guide keeps returning to that ratio as the decisive input.

The operational difference shows up on the automation meter. Journey depth is metered per run, so the more playbooks a SaaS runs — onboarding, dunning, winback, expansion, lifecycle broadcasts — the more the run meter compounds against it. Teams that expect their journey count to grow every quarter should price that trajectory explicitly rather than the current month.

Fairness requires stating where Resend keeps the advantage. Its transactional API discipline, per-message forensics, and engineering-native workflow are genuine strengths, and a team whose entire email program is receipts and notifications may never touch the marketing or automation meters at all. On that shape, Resend's transactional Pro at $20 for 50,000 is an excellent, simple answer, and lifecycle depth is an irrelevant criterion.

The honest way to hold both claims at once is to price them against your own mix rather than a category average. Take the last three months of transactional sends, stored contacts, and automation runs, price them under Resend's verified rates, then compare the result with a send-metered alternative and let the twelve-month projection, not the feature list, choose the vendor.

When running both tools is the right call

Some stacks pair Resend with a lifecycle system under marketer and agent governance. The arrangement buys separation of concerns — different blast radius, different approval culture, different audit trail — and its cost is two contracts, two suppression lists, and two places where unsubscribe state can drift. That cost is acceptable when the separation is deliberate; it is not acceptable when it is merely historical.

If you take the dual-vendor path, assign ownership explicitly: which system sends password resets, which sends dunning, and which is authoritative for suppression. Ambiguity there is the failure mode, not pricing — both bills are legible, but a subscriber unsubscribed in one system and still mailed by the other is a compliance problem no discount offsets.

When consolidation makes sense

The consolidation case strengthens as billing events become daily work, as the contact database outgrows engagement, and as the second contract starts requiring its own maintenance hours. At that point the question is not which tool sends better email but which system should know subscription state first — and a platform whose triggers start at the processor removes a synchronization step that a three-vendor stack can never fully retire.

• Choose Resend when engineers own email, transactional volume dominates, and the marketing plus automation meters stay small.
• Run both when the separation of transactional and lifecycle reputation is a deliberate architecture decision with named owners.
• Model all three Resend meters against the pooled alternative at twelve months, not at signing.
• Re-verify both vendors' live pricing pages before committing; this comparison is a September 2026 snapshot.

Resend's transactional clarity is the deciding advantage when engineers own delivery and the three-meter model maps cleanly to the architecture.

Resend questions and answers

What exactly does Resend charge for?

Three independent meters: transactional emails sent, marketing contacts stored, and automation runs consumed. Nothing is bundled across them, so each grows for its own reasons and appears on its own line.

Which meter usually surprises new teams?

Automation runs. Contacts and sends are visible in every dashboard, but runs accumulate per enrollment traversal, so a three-step flow with 4,000 monthly enrollees silently consumes 12,000 runs before any other workflow exists.

Is the free tier usable for a real product?

For transactional mail, yes: 3,000 emails per month is enough for early auth, receipts, and notifications, provided the 100-per-day cap does not collide with a launch spike.

How do overages behave?

Transactional overage is linear at $0.90 per additional thousand emails, and automation overage is linear at $0.0015 per additional run, so both are easy to forecast and easy to cap with alerts.

Do I need both transactional and marketing plans?

You need whichever meters your program actually uses. A receipts-only product can live entirely on the transactional side; a newsletter-only project on the contact side; most SaaS programs touch all three.

How does this compare with pooled pricing?

Pooled models charge one meter for everything, trading line-item granularity for forecasting simplicity.

One last reminder before the verdict: automation runs are metered per enrollment traversal rather than per campaign, so the practical forecasting unit is recipients multiplied by steps — and the practical safeguard is a run estimate attached to every workflow proposal, reviewed alongside sends and contacts in the same monthly pass.

Verdict and checklist

Choose Resend when engineers own email, transactional reliability is the top risk, and the three-meter structure maps cleanly to your architecture. Forecast all three meters, keep list hygiene tight on the marketing side, budget automation runs per playbook, and never launch on the free daily cap. Re-verify every figure at Resend pricing, then return to the pricing index for the full verified set.

A sensible rollout order

The sequence below keeps every meter observed before it can surprise anyone, and it works whether the account is a first production send or a migration from a single-pool vendor.

• Instrument all three meters on day one, even while only one of them is consuming anything.
• Prove transactional sending against real traffic before enabling marketing contacts or automation enrollment.
• Set alert thresholds at eighty percent of each allowance so overage becomes a scheduled decision.
• Clean and suppress the contact list before the first import lands on the marketing meter.
• Attach a projected run count to every new automation, the way engineers attach latency budgets to endpoints.
• Time the transactional Pro step ahead of known launch windows rather than during the spike itself.
• Run one human-approved broadcast end to end, including suppression and test-render checks, before scaling cadence.
• Reconcile the first real invoice against your forecast line by line and correct the model, not just the budget.

What to review every month

Four numbers are enough: transactional sends against the 50,000 Pro allowance, stored contacts against the current marketing tier, automation runs consumed against the 10,000 included plus actual overage, and revenue attributable to the sequences that produced them. Anything that moves twenty percent in a month deserves a reason before the next cycle begins.

Pair that review with a hygiene pass — suppress hard bounces, sunset chronic non-openers, archive dead workflows — because on this model every one of those actions is a direct pricing control as well as a deliverability one. Two hours of monthly maintenance routinely costs less than a single avoidable tier step.

When to reconsider the model itself

Reconsider the three-meter structure when the contact database grows faster than engagement for several quarters running, when automation run counts climb steadily despite workflow audits, or when the team spends more time forecasting meters than improving the program. Each of those is a signal that the geometry, not the volume, has become the constraint.

For the broader field, start from the pricing index, read the 15-tool comparison for positioning context, and take operating questions — approval thresholds, staging discipline, cost attribution per agent run — into the discussion threads, where thirty-eight scenarios pressure-test the governance side before it becomes an incident.

Whichever meter you pick, forecast three months of transactional sends, marketing contacts, and automation enrollments, keep suppression tight, gate consequential sends behind a human, and use Resend's current pricing page as the final authority when you pay.

Open pricing index