Mailgun Pricing 2026: Developer Email API, Volume Ladder

Mailgun (Sinch Email) is a developer-first sending API with one of the clearest volume ladders in infrastructure email. Verified September 2026: Free covers 100 emails per day with no expiration, Basic starts at $15/month (10,000 emails), Foundation at $35/month (50,000 emails), Scale at $90/month (100,000 emails), with overage from $1.80 per 1,000 emails and list-validations from $1.20 per 100. Optimize, Validate, and Inspect are separately priced products. Source: mailgun.com/pricing.

Who this guide is for

Mailgun is infrastructure, not a marketing suite. It fits teams that think in API calls and want the clearest volume ladder in the category.

The deciding question is whether email is infrastructure or growth for you today. Infrastructure buys Mailgun; growth eventually buys consolidation.

Whichever answer you give now, keep the dependency audit current — the cost of moving later is almost always the integrations nobody wrote down.

Pricing geometry explained

Mailgun meters delivered volume against fixed monthly bands: 10,000 emails at $15, 50,000 at $35, and 100,000 at $90, with the free tier holding at 100 emails per day indefinitely.

Two properties make this ladder unusually legible. Each step roughly multiplies capacity for a small price increase, and overshooting a band costs $1.80 per 1,000 extra emails rather than an immediate plan change — so growth shows up as dollars, not as forced migrations.

The counter-pressure is the volume-band jump itself: crossing from Foundation to Scale is a discrete step, and teams hovering just under a threshold pay either for headroom they do not use or for overage they did not plan.

Everything outside sending is a separate commercial conversation. List validation, placement optimization, and inbox testing are distinct products — Validate, Optimize, and Inspect — which means hygiene and testing budgets sit outside your send budget by design.

Plan with three numbers in hand: average monthly sends, burst peaks (launches, password-reset storms), and a validation cadence. Only the first maps to the band; the other two land as overage or as add-on spend.

Model with burst headroom rather than average headroom. Infrastructure email fails visibly on the day traffic spikes, which is exactly the day a band chosen on averages is too small.

Plan-tier ladder

1. Free100/day forever. Staging environments, MVPs, and low-volume alerts. One domain and 1-day log retention keep it honest.
2. Basic $1510k emails/month with monthly overage. First production tier for transactional mail, notifications, and small campaigns via API.
3. Foundation $3550k emails/month. Covers most growing SaaS transactional volume with room for lifecycle sends alongside.
4. Scale $90100k emails/month plus deliverability depth. Buy here when inbox placement becomes a managed discipline, not a hope.
5. EnterpriseCustom volume with bundled deliverability services and support. Negotiate log retention, IP strategy, and incident SLAs.

Worked cost shapes

Mailgun's unit economics are brutally favorable for transactional volume — $90 covers what costs marketing platforms hundreds. The catch is everything Mailgun does not do: no journeys, no segmentation UI, no billing triggers, no attribution.

Hidden fees and fine print

The send ladder is honest. The exposure lives in the products that surround it and in the step where a band stops being enough.

• Volume-band jumps: crossing 10k, 50k, or 100k emails is a discrete tier step — plan the threshold before launch week hits it.
• Overage rates: from $1.80 per 1,000 extra emails, cheap per unit but capable of exceeding a tier upgrade at sustained volume.
• List validation: Validate bills from $1.20 per 100 checks — hygiene is a separate budget line, not a plan feature.
• Optimize and Inspect: placement optimization and inbox testing are separately priced products.
• Compliance work sits with you: consent, suppression discipline, and blocklist monitoring are operational responsibilities rather than plan inclusions.
• Shared-pool risk: cheap entry inherits pool-mate reputation; a bad neighbor affects your placement.
• Dedicated IP management: warm-up, monitoring, and rotation are self-run unless you buy support around them.
• Domain structure: one domain on Free, five inbound routes on Basic — multi-domain architectures need higher tiers or consolidation.
• Log retention: one day on Free; investigation and incident forensics require a paid tier.
• Ticket support only at entry: response expectations change with tier, so match support needs to plan, not to price.
• Free-tier daily cap: 100 emails per day cannot absorb bursts regardless of how far below monthly volume you are.
• Enterprise bundling: dedicated support and bundled tooling live behind custom pricing, so late-stage negotiation sets the number.

Ask support to confirm overage rates, validation pricing, and retention in writing before purchase. Every line above is public; the surprises come from budgeting sends alone while running a deliverability practice.

The structural point: Mailgun prices the pipe, not the program. Compliance, hygiene, and testing are real costs with real owners — they simply live in different columns of the budget.

Cost scenarios

These three monthly examples translate Mailgun's volume ladder into operating scenarios. Recheck the current band and overage rate at purchase time, then use the pricing index to compare the resulting send cost with the other API providers.

Rows two and three show the ladder working as designed: the step from $15 to $35 buys five times the volume, which is why overshoots are usually cheaper to absorb than to pre-purchase.

Row four is where the surrounding products enter the budget. At 120,000 sends, validation runs on a cadence and placement work stops being optional — both land outside the $90.

Compare that all-in figure against marketing platforms before assuming infrastructure is always cheaper. It is cheaper for sending; the question is whether sending is the whole job.

One caution on overage: it is cheap per unit precisely because it is expected to be temporary. If overage appears three months running, the band is wrong and the tier step is cheaper than the habit.

Cost-control playbook

Volume-priced infrastructure rewards engineering discipline rather than list hygiene — different habits, same principle: review the meters on a schedule.

• Track sends per day and per month separately; the free tier binds daily while paid tiers bind monthly.
• Forecast burst events (launches, resets, incidents) against band headroom before they occur.
• Run validation on a fixed cadence and count it in the monthly total from day one.
• Compare overage spend against the next tier's price each quarter — the crossover decides the jump.
• Catalog every sending domain, API key, webhook, and route so tier changes never break an unknown dependency.
• Keep suppression data portable; blocklist incidents are cheaper to survive with exports ready.
• Separate transactional and marketing subdomains deliberately if you ever add marketing volume.
• Review log retention against your incident-response needs; forensics on a one-day window is a myth.
• Re-price Optimize, Validate, and Inspect usage annually — tooling spend drifts like send spend.
• Assign an owner for IP and blocklist watch if you run dedicated infrastructure.
• Compare cost per delivered email against consolidation alternatives once lifecycle work appears.
• Re-check the ladder at renewal: volume bands are the part of this market that moves.

Teams that review volume monthly step between tiers on purpose. Teams that review annually discover the step on an invoice that already cleared.

Enterprise fit

Mailgun fits engineering teams that treat email as an API call: transactional receipts, notifications, and alerts with RESTful APIs, SMTP relay, tracking, webhooks, and inbound routing. The always-on free tier, transparent overage, and separate deliverability products (Optimize for placement, Validate for hygiene, Inspect for testing) make it the most legible infrastructure choice for developers.

It does not fit marketing lifecycle. There is no journey builder, no audience segmentation surface, no AI drafting, and no notion of a trial, a plan, or a failed payment. Teams that start with Mailgun for receipts and then need growth email face a build-vs-buy fork: bolt on a marketing platform (two vendors, two reputations) or consolidate lifecycle onto one system.

Deliverability responsibility also sits heavier on Mailgun operators than on managed marketing suites. Shared-pool senders inherit pool-mate risk, while dedicated-IP tiers demand self-run warm-up and monitoring discipline — the platform provides the knobs, not the hands. Budget accordingly: either assign engineering or deliverability time to IP management and blocklist watch, or step up to Scale and Enterprise where placement support and specialist guidance absorb that load.

Migration checklist (complement or consolidate)

• Catalog every Mailgun dependency: sending domains, API keys, webhook endpoints, inbound routes, and suppression data.
• If keeping Mailgun for transactional, isolate reputations deliberately: separate subdomains for transactional vs marketing with distinct authentication.
• If consolidating, migrate webhooks first — bounce and complaint streams feed suppression and dunning logic downstream.
• Validate list hygiene with Validate (or equivalent) before any bulk migration; infrastructure IPs amplify bad lists fast.
• Load-test burst scenarios (password-reset storms, launch days) against the new platform before DNS cutover.

AI scope and the lifecycle boundary

Mailgun's pricing page describes an email API, SMTP relay, inbound routing, validations, dedicated IPs, and deliverability analytics. It does not list an AI copywriter, AI segmentation assistant, AI journey builder, or AI send-time optimizer as an included plan feature. Mailgun can deliver a message generated by your own application, but its pricing page does not promise to generate the message or decide who receives it.

If an AI workflow is part of your brief, separate the jobs. Your application or a model service can draft a subject line, classify an event, or choose a template; Mailgun can accept the rendered message and report delivery events. The model call, prompt logging, review process, and data controls sit outside the Mailgun subscription. For any Mailgun AI add-on not named on the pricing page, pricing on request is the only responsible budget line.