Sendly vs Mailgun

The short version

Mailgun is an email API with a deliverability toolkit around it — validation, routing, logs. Teams that chose it usually did so because they wanted the API first and the dashboard second.

Sendly is built the same way round: everything the dashboard does is in a published REST contract, and both official SDKs are generated from it. What it adds is the half most transactional providers leave to a second vendor — marketing campaigns, a visual workflow builder, dynamic segments, and mailboxes that receive real mail on your domain.

What it does not add is a validation product or a separate inbound routing DSL. If those are load-bearing for you, check that against the list below before you plan anything.

mailgun.js keeps working, with a URL override

Sendly speaks Mailgun's transactional-send dialect on a dedicated endpoint at /api/compat/mailgun. Keep the Mailgun SDK you already run, point it at Sendly, and swap the key. There is no rewrite and no second client library.

The endpoint is a translator, not a side door. It parses the vendor request shape, maps it onto Sendly's native send, and runs it through the same pipeline your native API calls use — so verified-domain checks, sending limits, suppression and billing all still apply.

Where a dialect can express something Sendly does not support in this phase, the endpoint returns a clean error in that vendor's own error shape rather than dropping the field silently. Your existing error handling fires exactly as it would have.

Sendly implements Mailgun's POST /v3/<domain>/messages. Authentication is Mailgun's own HTTP Basic style, where the username is the literal string api and the password is your Sendly secret key, so the SDK's client options are the only thing that changes.

import formData from "form-data";
import Mailgun from "mailgun.js";

const mailgun = new Mailgun(formData);
const mg = mailgun.client({
  username: "api",
  key: process.env.SENDLY_API_KEY,
  url: "https://api.sendly.now/api/compat/mailgun",
});

// unchanged from here down
await mg.messages.create("yourdomain.com", {
  from: "you@yourdomain.com",
  to: "customer@example.com",
  subject: "Your receipt",
  html: "<p>Thanks for your order.</p>",
});

Two gotchas worth reading before you switch a production sender

• The domain in the URL path must match the domain of your from address. A mismatch is a 400 in Mailgun's error shape, not a best-effort send.
• Mailgun options past o:tag — scheduled delivery via o:deliverytime, tracking toggles, stored templates — are not applied. Unlike the Resend and SendGrid dialects, which reject what they cannot map, the Mailgun dialect honours the fields in the supported list and ignores the rest. If your sends lean on those options, that is a behaviour change to test for, not a footnote.

Mailgun publishes no official Python SDK, so Python senders either call the compat endpoint over plain multipart HTTP or use the Sendly Python SDK directly — for a Python codebase, moving to the native SDK is usually less work than preserving the vendor shape.

What Sendly does

Sendly is one workspace for transactional email, marketing campaigns and automation, against one contact database. Everything below is in the product today, not on a roadmap.

What Sendly costs

There is no monthly platform fee and there is no plan to outgrow. A free tier, metered usage that gets cheaper as volume rises, and a conversation for anything past that.

The free tier is a real tier, not a trial: it does not expire, and nothing about it is time-limited. Its allowance is shared across transactional sends, campaigns, workflows and inbound mail, because Sendly meters emails rather than products.

What to check on Mailgun's side

We do not publish Mailgun's prices or limits here. They change without telling us, and a stale number on this page would be the one thing on it you could catch us being wrong about. These are the questions worth answering from their own documentation before you move anything.

• What does the bill look like at your actual monthly volume, and at ten times it? Flat plans and metered curves cross over somewhere; find where.
• Are transactional sends and marketing sends billed and rate-limited from the same allowance, or two?
• Is there a contact-count charge on top of the send charge, and does an unsubscribed contact still count?
• What happens on the day you exceed the plan — a hard stop, an overage rate, or an upgrade prompt?
• Can you export your contacts, templates and suppression list in full, without asking support?
• Is there an API for everything the dashboard can do, and is the contract published?

Questions people ask

Does Sendly do email address validation?

Not as a standalone product. Sendly maintains a suppression list, handles bounces and complaints, and will stop you mailing an address that asked you to stop — but there is no pre-send verification API to check whether an address exists. If you buy Mailgun for validation specifically, that part does not have an equivalent here.

Can I receive mail on Sendly the way Mailgun routes handle inbound?

Sendly gives you real mailboxes on your own domain, reachable over IMAP and SMTP, and projects the messages into the API so your code can read them. It is a mailbox model rather than a routing-rule model: there is no expression language that forwards or posts a matching message onward. Read the receiving-emails guide before assuming a route maps across.

Will my o:tag values survive?

They map onto Sendly tags, sanitized. It is a partial mapping rather than an identity, so once you have moved, treat the tag values Sendly reports as the source of truth rather than assuming byte equality with what you sent.

What does a bad key look like?

A 401 with Mailgun's bare { message } error body, so mailgun.js raises the error your existing handling already catches and error.status and error.details read the way they do today.