Mailgun Pricing Guide for 2026

Who this guide is for

This guide is for the infrastructure buyer: platform and backend teams who treat email as a subsystem with SLAs rather than as a marketing channel, developers integrating REST APIs and SMTP relay under scoped keys, and operations leads who need validation, log retention, and reputation controls named as procurement requirements. It assumes you are comfortable choosing between stepped volume tiers because the binding constraints here are operational — domains, keys, logs, IPs — as much as volumetric.

It also serves the team pricing a two-vendor stack rather than a single platform. Mailgun covers the sending pipe while lifecycle playbooks live elsewhere, and this guide's job is to make the pipe's cost legible: stepped tiers from $15 per 10,000 through $35 per 50,000 and $90 per 100,000, overages from $1.80 per 1,000, validation meters billed separately, and the tier edges where retention, domains, and dedicated IPs change the production price. All figures were verified on the official page on September 17, 2026.

Mailgun's numbers below are presented neutrally so the infrastructure-versus-lifecycle tradeoff can be judged on its own terms. For the whole field first, start at the pricing index.

Read this guide closely if

• Engineers own email delivery and want programmable control over sending, validation, and suppression.
• Multi-tenant or multi-brand sending means domain and route limits are real constraints, not trivia.
• Log retention, webhook design, and reputation tooling appear in your procurement checklist.
• Volume is predictable enough to pick a tier and irregular enough that overage rates matter.
• List cleaning is a recurring program rather than a one-time migration chore.
• You are explicitly evaluating a two-vendor stack: infrastructure pipe plus lifecycle platform.

Look elsewhere first if

• A visual journey canvas, revenue attribution, or billing-triggered flows are required on day one.
• One bundled invoice with no separate validation line is a finance requirement.
• Your team has no deliverability owner, since this platform exposes levers rather than hiding them.
• Enterprise-grade dedicated infrastructure and burst SLAs are needed immediately rather than eventually.

Those exclusions describe fit rather than quality. Mailgun is a serious sending engine whose value increases with the expertise pointed at it; teams without that expertise will get more from a platform that hides the levers, which is a reasonable trade rather than a compromise.

The billing model in one paragraph

Mailgun charges for monthly sending volume in stepped plans with optional overages, keeping the classic infrastructure shape: pick the tier nearest your volume, burst above it when needed, and pay per-thousand overage rates rather than upgrading mid-spike. Email validation is a separate meter, so list-cleaning costs stay visible instead of hiding inside the send allowance. Deliverability services such as dedicated IPs, inbox placement testing, and managed reputation arrive as higher-tier features or add-ons, which means the sticker price and the production-hardened price can differ meaningfully.

This separation of send, validate, and optimize into distinct products mirrors how platform teams actually buy email. Development owns the send API, growth owns validation hygiene, and deliverability specialists own reputation tooling, each with its own budget line. Agent-run programs should mirror that split in their own cost attribution, tagging sends, validations, and reputation services separately rather than blending them into one email line item.

Pricing geometry explained

Mailgun's geometry is stepped volume with optional overages: choose the tier nearest your monthly sending, burst above it when reality disagrees, and pay per-thousand rates rather than upgrading mid-cycle. Alongside that sits a second, independent meter for email validation, which keeps list-cleaning costs visible instead of burying them inside the send allowance. Two meters, one step ladder, and a rate card for the space between the steps.

The stepped structure has a predictable consequence for growing programs: the bill moves in plateaus, not smoothly. A month inside a tier costs the same as a month half-empty within it, and the first send above the tier costs $1.80 per thousand rather than a new subscription. That shape rewards accurate sizing and forgives estimation error, which is exactly what infrastructure buyers want when traffic is spiky.

Validation metered separately is the more philosophically interesting choice. It prices hygiene as its own activity — and at the verified rate of $1.20 per 100 validations, every cleaning program scales with record count rather than send volume — so a team can spend deliberately on list quality without that spend being mistaken for sending volume. Cost attribution gets cleaner: sends, validations, and reputation services each carry their own line.

For a growing list, the geometry is neutral on storage and active on activity. There is no contact meter anywhere in the verified figures, so a database that doubles overnight changes nothing until it is mailed — and when it is mailed in bulk, the relevant question is which tier's plateau you now sit in rather than which contact band you crossed.

What a growing list does to the bill

Model monthly sends first and let the tier ladder follow. A program growing steadily from 8,000 to 12,000 to 18,000 emails crosses from free into Basic territory and then sits comfortably inside Basic's 10,000-allowance-plus-overage arithmetic — at $15 plus 8,000 overage emails at $1.80 per thousand, the overshoot costs about fourteen dollars and forty cents rather than forcing a tier change. The plateau structure turns growth into small, forecastable line items.

The second curve to track is validation volume. If list cleaning runs monthly against a growing database, validation spend scales with the record count rather than the send count, and the two curves should be budgeted separately. Teams that validate only before migrations discover a lumpy invoice; teams that validate on a schedule discover a modest recurring line they can forecast like any other utility.

The two rows without sending volume are the ones infrastructure teams should stare at. Validation is a metered product with real recurring cost, and log retention is a capability whose absence cannot be remedied by paying overage — both must be planned at tier-selection time rather than discovered later.

Everything else in the table is arithmetic you can redo from the verified anchors: tier prices, overage rates, and validation rates were confirmed on the official page on September 17, 2026, and the shapes above deliberately avoid any figure our sources did not render.

One final geometry note for forecasters: because both meters are linear between steps, a spreadsheet model of this platform needs only tier thresholds and two rate cards — no feature matrix, no seat logic, no contact bands. The discipline lives in getting the two input curves right rather than in decoding the pricing structure.

Plan tiers, verified

The free plan is a genuine integration sandbox: 100 emails per day, RESTful APIs and SMTP relay, one custom sending domain, two API keys, tracking with analytics and webhooks, one inbound route, and one-day log retention with ticket support. A developer can integrate, test webhooks, and run staging traffic indefinitely without a card. The single-domain limit and one-day logs are the real constraints: multi-tenant staging and post-incident forensics both outgrow free quickly.

Basic starting at $15 per month for 10,000 included emails removes the daily limit, keeps one domain and two keys, extends to five inbound routes, and opens monthly overage options with extra emails from $1.80 per 1,000 and extra validations from $1.20 per 100. Foundation at $35 per month for 50,000 emails is the step where production programs land, adding the headroom and reputation features a growing SaaS needs. Scale at $90 per month for 100,000 emails serves high-volume senders with the deliverability posture to match, and Enterprise packaging beyond that tailors dedicated infrastructure with burst-sending SLAs for senders whose normal Tuesday exceeds other tools' monthly records.

Limits, overages, and the fine print

Overage math on Mailgun stays gentle by design. Extra emails from $1.80 per 1,000 mean a 20,000-send overshoot month costs roughly thirty-six dollars above the tier, so launch spikes and seasonal bursts translate to tens of dollars rather than forced mid-cycle upgrades. Validations from $1.20 per 100 keep list-cleaning honest: scrubbing a 50,000-record legacy import costs on the order of sixty dollars, a one-time hygiene investment that pays back in deliverability every month after.

The limits that actually bite are operational rather than volumetric. One-day log retention on free and Basic means debugging a deliverability complaint from last week is impossible without an upgraded plan or an external log sink, so teams with compliance or support-ticket workflows should price retention as carefully as sends. Single-domain sending on lower tiers forces multi-brand teams upward, and dedicated IPs only enter the picture at serious volume, so shared-IP reputation dynamics apply to everyone below that line.

Hidden fees and fine print

Mailgun's surprises are product-shaped rather than contractual: separate meters, capability ceilings, and tier edges that change what production costs. The list below is ordered by how often each one reshapes a budget after teams have already migrated.

Validation is a second bill

Email validation is metered separately from sending, with extra validations from $1.20 per 100 — a recurring cost for teams that validate on a schedule, sized in blocks of a hundred records rather than in sends. Because the meter is invisible in send-focused dashboards, monthly hygiene programs quietly accumulate a line item that nobody attached to a budget line.

The cure is attribution rather than avoidance: tag validation spend as its own category, forecast it against database size rather than send volume, and treat the sixty-dollar import scrub as the cheapest deliverability investment available — it pays back every month after in reputation alone.

Volume jumps and the plateau edge

The tier ladder — $15 for 10,000, $35 for 50,000, $90 for 100,000 — steps in large increments, so a program sitting just above a tier boundary pays overage at $1.80 per thousand rather than jumping to the next plateau. That is fine for occasional overshoots and expensive as a permanent arrangement: a sustained 20,000-send month on Basic costs $15 plus about eighteen dollars of overage, money that should instead trigger a sizing conversation.

Model peak months rather than averages when choosing tiers, and treat three consecutive months of overage as a signal to re-ladder rather than a routine to accept. The same discipline applies above Scale, where packaging becomes a vendor conversation rather than a published rate.

Domains, keys, and routes

Domain and key limits bind earlier than volume does. Free and Basic carry one custom sending domain and two API keys, five inbound routes arrive at Basic, and multi-brand or multi-tenant architectures hit those ceilings before they hit send allowances — at which point the effective cost of the tier is the migration effort to the next one.

Confirm the exact per-tier entitlements on the live page, since these are the capabilities that move between tiers more often than headline prices. Teams running several client brands should map domains to tiers before committing rather than discovering the constraint during onboarding.

Log retention and operational capability

One-day log retention on free and Basic means last week's deliverability complaint cannot be investigated without an upgraded plan or an external log sink. That is an operational cost disguised as a storage setting: teams with support workflows or compliance expectations should price retention as carefully as sends, and sink logs externally if the tier arithmetic stings.

Dedicated IPs and advanced reputation tooling enter the picture at higher tiers, which means shared-IP reputation dynamics apply to everyone below that line. Factor that in when forecasting: on lower tiers, your neighbors' behavior is partly your deliverability, and no overage payment resolves it.

A pre-signing checklist

• Forecast monthly sends and validation volume as two separate curves, then size the tier to both.
• Treat three consecutive overage months as a signal to re-ladder rather than a routine to accept.
• Map required domains, keys, and inbound routes against each tier's entitlements before purchase.
• Decide whether logs sink externally or whether retention tiers must be budgeted from day one.
• Confirm shared versus dedicated IP expectations for your volume band and reputation requirements.
• Check Enterprise and burst-SLA terms live if any month could plausibly exceed Scale's 100,000.
• Re-verify every figure at Mailgun pricing on decision day — all figures here were verified September 17, 2026.

What the free tier really proves

One hundred emails per day covers integration testing, staging environments, side projects, and the earliest transaction flows: verification codes, receipts, and founder-sent onboarding to a handful of daily signups. Because the API surface on free matches production, code written against the sandbox ships unchanged, which is exactly what an infrastructure free tier should guarantee. Webhook handling, bounce processing, and inbound routes can all be proven before spending.

The honest boundary is anything resembling production operations. One-day logs, one domain, and ticket-only support mean the first real incident, the second brand, or the first compliance question each force an upgrade conversation. Budget Basic at $15 as the true starting line for production, and treat free as the permanent staging tier it is designed to be.

Cost scenarios: three monthly invoices

These three months use only figures verified in this guide and the pricing index: free at 100 emails per day, Basic at $15 for 10,000, Foundation at $35 for 50,000, Scale at $90 for 100,000, overages from $1.80 per 1,000, and validations from $1.20 per 100. Validation spend is shown as blocks of a hundred records so every figure can be checked rather than trusted.

Month two: sandbox and first traffic

The integration is live and the product sends about 2,100 emails across verification codes, receipts, and a founder-led onboarding sequence — comfortably inside the 100-per-day free ceiling. Webhooks are proven, bounce processing is observed, and the invoice is zero dollars.

The month's real output is confidence: code written against the free tier's full API surface ships unchanged to production, which is exactly what an infrastructure free tier should guarantee. The constraints worth noting are the single sending domain, two API keys, and one-day log retention, none of which bite until operations rather than integration become the daily work.

Month five: production entry with hygiene

Sending grows to 18,000 emails, overshooting Basic's 10,000 allowance by 8,000, while a scheduled hygiene pass validates 2,000 records. The invoice is Basic at $15, plus 8,000 overage emails at $1.80 per thousand for about fourteen dollars and forty cents, plus twenty validation blocks at $1.20 for twenty-four dollars.

Totaling roughly fifty-three dollars, the month shows the two-meter design working as intended: the send overshoot is small and linear, the validation line is visible and attributable to a deliberate hygiene program, and neither is hidden inside a bundled tier. The decision the month raises is whether the overage becomes routine — if so, Foundation at $35 for 50,000 is the next plateau to model.

Month twelve: high-volume steady state

The program now sends 95,000 emails monthly, inside Scale at $90 for 100,000, with a recurring validation sweep across 10,000 records — one hundred blocks at $1.20, or a hundred and twenty dollars. The invoice is about two hundred ten dollars, of which more than half is validation rather than sending.

That ratio is the scenario's lesson: at scale, hygiene can outweigh volume on an infrastructure bill, and teams that only forecast sends will be perpetually surprised by the larger line. Budgeting both curves together — the discipline this guide asks for in the geometry section — turns the invoice into two predictable utilities instead of one mystery.

Read the validation row across all three months: it starts at zero, appears as deliberate hygiene, then becomes the majority line. No contact meter exists anywhere in the model, so none of these totals reacts to how large the stored database grows — only to what is sent and what is cleaned.

Cross-referenced against the wider verified set — Resend's transactional Pro at $20 for 50,000 and Postmark's Basic at $15.00 per 10,000 — Mailgun's shape reads as infrastructure with a cleaning bill attached, which is precisely the value proposition for teams that already know they need both.

Who Mailgun is for

Mailgun fits developers and platform teams that own email as infrastructure: SaaS products sending receipts, notifications, and invites at volume; marketplaces with per-tenant sending needs; and fintech or healthcare-adjacent products where log retention, validation, and reputation controls are procurement requirements. Teams with in-house deliverability knowledge extract the most value, since the platform exposes levers rather than hiding them.

Mailgun is a weaker fit for marketers seeking lifecycle playbooks. There are no billing-triggered dunning flows, no visual journey canvas for growth teams, and no revenue attribution out of the box. Those programs need a lifecycle layer above the pipe, which is why many stacks pair Mailgun-style infrastructure with a lifecycle platform, accepting two bills for best-of-breed fit.

Agent and MCP fit

Agents thrive on Mailgun's API-first surface: sending, validation, suppression management, and event webhooks are all machine-shaped primitives that script cleanly under scoped API keys. An agent can validate a list before a campaign, monitor bounce webhooks for reputation anomalies, rotate keys on schedule, and draft the suppression analysis a human then approves. Inbound routes even let agents process replies programmatically, closing the loop from send to response.

Governance must be built, not assumed. There is no native MCP server or approval queue, so teams wrap the API with their own proposal-and-approve layer: agents get send rights to staging domains and propose-only rights on production, rate limits cap blast radius, and every production send carries human sign-off per our discussion scenarios on thresholds and staging discipline. Log retention tiers double as the audit trail budget, so retention upgrades are governance spending, not overhead.

How Mailgun compares with lifecycle platforms

Mailgun's stepped sending and validation meters suit teams that want direct control of the delivery pipe. Lifecycle platforms add billing events, journey orchestration, and attribution, so SaaS buyers should price processor integration and suppression synchronization beside the subscription rather than treating the glue as free.

The honest counter-case is expertise. A team with real deliverability knowledge extracts more from Mailgun's exposed levers than from any bundled lifecycle tool, and pairing Mailgun for transactional volume with a lifecycle platform for campaigns remains a defensible two-vendor architecture — provided the stack is priced as two contracts, two suppression lists, and two warm-up histories rather than as a free optimization.

Where each one is the clear answer

Mailgun is the clear answer when the job is programmable delivery: per-tenant sending, validation pipelines, inbound routing, webhook-driven suppression, and reputation tooling under direct control. On that shape, journey depth is irrelevant, and the validation meter is a feature rather than a nuisance because hygiene is an engineering program.

A fair decision sequence

• List every workflow that must start from a billing event and count how many need custom integration on infrastructure-only tooling.
• Forecast two curves — monthly sends and validation volume — then price them across Mailgun's verified steps.
• Add the build-and-maintain cost of processor webhooks and suppression sync to the infrastructure total.
• Add log retention and domain requirements to the infrastructure total where support or compliance demands them.
• Before choosing Mailgun, confirm the current volume and validation meters against your forecast; this guide is a September 2026 snapshot.

The structure above names each vendor's verified anchors and architectural gaps plainly, leaving the choice to your integration roadmap and your team's deliverability depth.

Mailgun questions and answers

What does Mailgun charge for?

Monthly sending volume on stepped plans, plus separately metered email validation. The verified anchors are free 100 emails per day, Basic at $15 for 10,000, Foundation at $35 for 50,000, Scale at $90 for 100,000, overages from $1.80 per 1,000, and validations from $1.20 per 100.

What does the free tier really allow?

One hundred emails per day with full API access, one custom sending domain, two API keys, tracking and webhooks, one inbound route, one-day log retention, and ticket support — an integration sandbox rather than a trial countdown.

How do overages work?

Extra emails bill from $1.80 per 1,000, so a 10,000-email overshoot adds about eighteen dollars rather than forcing a mid-cycle tier change — linear, predictable, and worth alerting against.

Why is validation billed separately?

Because list hygiene is a distinct engineering activity with its own volume driver — record count rather than send count — and pricing it separately keeps hygiene spend visible in cost attribution.

What are the non-volume constraints?

Domains, API keys, inbound routes, and log retention all vary by tier, and on lower tiers they bind before send volume does; one-day retention in particular makes last week's deliverability complaint impossible to investigate without an upgrade or external sink.

How does this compare with pooled per-email pricing?

Verdict and checklist

Choose Mailgun when engineers own email, programmatic control and validation matter more than playbooks, and volume justifies infrastructure thinking. Start free for integration, enter production at Basic $15 per 10,000, step to Foundation $35 per 50,000 or Scale $90 per 100,000 as volume dictates, budget overages from $1.80 per thousand and validations from $1.20 per hundred, and sink logs externally if retention tiers sting. Use the dated Mailgun pricing page for the final check, then return to the wider pricing index.

A thirty-day infrastructure plan

The sequence below proves the pipe before production depends on it, and it works for a fresh integration or a migration from another sending API.

• Day one: connect the domain, rotate scoped API keys, and stand up a weekly sends-and-validations dashboard.
• Week one: prove webhook delivery, bounce processing, and suppression behavior against real traffic.
• Week one: record domain, key, route, and log-retention entitlements for your chosen tier.
• Week two: run one production-shaped send and observe pacing, queueing, and reputation signals.
• Week two: schedule the first validation sweep and record its block count as a baseline.
• Week three: test log availability against a real support question — can you answer it with one day of retention?
• Week three: compare the current tier plus overage against the next plateau using three months of actual volume.
• Week four: sink logs externally or confirm retention tiers in the budget, deliberately rather than by default.
• Week four: confirm dedicated-IP expectations for your volume band before reputation becomes the constraint.
• Day thirty: price the lifecycle layer you will need beside this pipe, since sending alone rarely completes a SaaS program.
• Day thirty: re-read the live pricing page and commit to today's figures rather than this guide's snapshot.

Signals to re-ladder or consolidate

Re-ladder when overage repeats three months running, when a second brand forces a domain-tier conversation, or when support workflows make one-day logs operationally untenable. Each is a sizing question that the verified rate card answers directly, without migration drama.

Consider consolidation when billing-triggered playbooks become daily work, when suppression state has to stay synchronized across two systems, or when the second contract's maintenance hours start rivaling the subscription itself. At that point the question shifts from which tool sends better email to which system should know subscription state first.

Neither move is urgent, and neither invalidates time spent on infrastructure: the pipe you built remains legible either way, and a stack that documents its own suppression, logs, and tier logic is easier to change than one held together by habit.

Where to go next

Start at the pricing index to place Mailgun's steps beside every other verified vendor, then read the 15-tool comparison for positioning context.

Mailgun's rate card cannot decide who may release a production send or how validation spend is assigned. Use the editorial discussion threads to pressure-test those approval and attribution rules before the first incident.

Pair this guide with Resend or Postmark when you are comparing transactional APIs, and inspect SendGrid for multi-tenant IP controls.

The closing discipline does not change with vendor: forecast your own sends and validations separately, keep suppression tight, gate consequential mail behind a human, and let the live pricing page overrule this September 17, 2026 snapshot at the moment you actually pay.

Open pricing index