Email for AI Agents: Send, Receive, and Reply Without an SMTP Rabbit Hole

Send and receive under one API key

The minimal email setup for an AI agent is a send API. It works — until the customer replies. At that point you need inbound parsing, threading, and a way to route the reply back to the agent that sent the original message. That is the gap a plain SMTP relay cannot fill:

import { createClient } from "@mailsai/sdk";
const mails = createClient({ apiKey: process.env.MAILS_API_KEY });

// Sending: one call, no DNS to configure, no domain to verify upfront.
await mails.messages.send({
  from: "hello",             // the agent every workspace gets at signup
  to: "customer@example.com",
  subject: "Your invoice #441 is ready",
  body_text: "Invoice #441 for August is attached — reply with any questions.",
});
// → sent from hello.yourworkspace@send.mails.ai, SPF/DKIM/DMARC aligned

// Receiving: replies arrive as typed events, already parsed and injection-scanned.
mails.agent("hello").onReply(async (event) => {
  // event.data.extracted_text — the reply's plain text, quoted history stripped
  // event.thread_id  — stable across the whole conversation
  // event.injection_score — 0.0–1.0; quarantined from 0.95
  // event.source_message_id — the received message, to reply to
  // Hold for a person: quarantined, not scanned (no score), or 0.5 and up
  if (event.quarantined || typeof event.injection_score !== "number" || event.injection_score >= 0.5) {
    return flagForReview(event);
  }
  if (!event.source_message_id) return;

  // Reply in the same thread — no Message-ID wrangling.
  await mails.messages.reply(event.source_message_id, {
    body_text: "Got it — I'll process that and follow up by end of day.",
  });
});

Reputation stays per agent, not per account

A shared sending domain fails when one sender misbehaves and the whole domain takes the bounce or complaint hit. Here, each AI email agent has its own reputation score. An agent whose complaint rate drifts above 0.3 % — below the upstream threshold of 0.5 % — is suspended automatically before the damage propagates. Your other agents keep sending unless the workspace’s mail as a whole crosses the same line.

For consistent high volume, around 300,000 messages a month, a dedicated IP is available on request on the Scale plan. It warms with your real mail, not a manual warmup schedule: see automated dedicated IP warm-up.

Custom domain, same model

By default agents send from an address on the shared sending domain, such as hello.yourworkspace@send.mails.ai — working email in one API call, no DNS required. When your agents need to carry your brand, connect a custom domain and every agent on it sends as agent@yourcompany.com, authenticated end to end. Custom domain email for AI agents covers the DNS steps — they’re generated for you and verified automatically.

The email API: one key, both directions

The email API for AI agents wraps sending, receiving, and classification into a single integration. You authenticate once with an API key, create agents programmatically, and every reply arrives as a structured event — no IMAP poll loop, no MIME parser, no OAuth refresh token. The API ships with TypeScript and Python SDKs, an MCP server for agent IDE integration, and a plain REST interface for everything else. The email API for AI agents solution page has code examples in both languages, or read the full email API comparison for 2026 to see how Mails.ai stacks up against Resend, SendGrid, Postmark, and Mailgun for agent use cases.

What it costs

The free tier covers 3,000 emails and 3,000 inbound replies a month with one agent — no card required. Beyond that, Pro is $20 a month for 50,000 emails and 50,000 inbound replies across five agents, and Scale is $99 for 250,000 emails and 500,000 inbound replies with unlimited agents. Both include the injection scan on every inbound and opt-in intent/entity classification of first-contact mail at no extra charge, and yearly billing is two months free. No per-seat fee. For the full management layer — routing, classification, and per-agent reputation — see AI email management.