# adbtd > adbtd is whitemark email infrastructure as a service. We provision a pristine, dedicated mail stack (Postfix + Dovecot + Rspamd, dedicated outbound IP, SPF/DKIM/DMARC aligned to your domain) under your own domain in minutes. Recipients see your domain in every header — no `via sender.io`, no third-party DKIM signer, no shared sending reputation. ## What makes adbtd different - **Whitemark by default**: SPF, DKIM, return-path, MX all aligned to the customer domain. No `via` lines, no provider domain in `From` or `Return-Path`. - **Dedicated IP, never shared, never recycled**: every account gets at least one fresh IP. No shared pool. No neighbours damaging your reputation. - **Two prices, no tiers**: $1.99 / € / £ per mailbox per month. $19 / € / £ per dedicated IP per month. Minimum 1 mailbox + 1 IP. Billed monthly. Cancel anytime. - **Live in under 10 minutes**: signup → order wizard → automated provisioning → SMTP/IMAP credentials in your dashboard. - **Use case-agnostic**: cold outreach, transactional, list ops — the IP is yours, the rules are yours. We provision and monitor reputation; we don't gate sends. ## Pricing - **Mailbox**: $1.99 / € 1.99 / £ 1.99 per mailbox per month, billed monthly. Includes IMAP + SMTP credentials, full ESMTP, no per-send pricing. - **Dedicated IP**: $19 / € 19 / £ 19 per IP per month. Every account ships with at least one. Add more, assign per domain, reassign anytime. - **Example**: 1 mailbox + 1 IP = $20.99 / mo. 240 mailboxes + 3 IPs = $534.60 / mo. - **Currency**: pick USD / EUR / GBP at first checkout. Locked on the workspace afterwards. ## Key URLs - Home: https://adbtd.com/ - Pricing: https://adbtd.com/pricing - Docs: https://adbtd.com/docs - API reference: https://adbtd.com/docs/api - Comparison vs Mailgun/Postmark/SendGrid/Amazon SES/Resend: https://adbtd.com/compare - 2026 provider comparison (adbtd, Infraforge, Mailforge, Maildoso, InboxKit, Mailgun, SendGrid, Postmark, SES, Resend): https://adbtd.com/blog/best-dedicated-smtp-infrastructure-2026 - Status: https://adbtd.com/status - Sign up: https://adbtd.com/signup - Log in: https://adbtd.com/login - Privacy: https://adbtd.com/privacy - Terms: https://adbtd.com/terms - Imprint: https://adbtd.com/imprint - Contact: hi@adbtd.com ## MCP (Model Context Protocol) integration adbtd exposes a remote MCP server at `https://mcp.adbtd.com` so AI agents (Claude Code, Claude Desktop, Cursor, or any MCP-compatible client) can manage the email stack programmatically. Authorization is OAuth 2.1 with Dynamic Client Registration; the first connect opens the browser to mint an API key under your workspace, and the connection is revocable from the dashboard at any time. The MCP toolset mirrors the REST API one-for-one — anything you can do via `/api/v1/*` you can do via MCP. ### MCP tools (workspace-scoped) - `get_me` — workspace info + subscription state. - `list_mailboxes` — list mailboxes, optionally filtered by domain or status. - `get_mailbox` — fetch one mailbox by id. - `create_mailbox` — provision a mailbox on a verified domain; returns SMTP/IMAP creds. - `delete_mailbox` — hard-delete a mailbox by id; irreversible. - `send_email` — send mail through a mailbox via SMTP submission. - `list_messages` — fetch recent messages from a mailbox over IMAP, paginated. - `get_message` — fetch full body + attachments by IMAP UID. - `list_domains` — list workspace domains with SPF/DKIM/DMARC status. - `add_domain` — add a BYO domain; DNS records to publish are returned in `list_domains`. - `delete_domain` — remove a domain and every mailbox on it; irreversible. - `list_ips` — list dedicated IPs and RBL listings. - `add_ips` — order N additional dedicated IPs; may return a Stripe checkout URL. - `release_ip` — release a dedicated IP; releasing the last IP cancels the subscription. - `get_deliverability` — 7-day sent volume, avg spam score, top symbols, RBL status. Setup docs: https://adbtd.com/mcp. ## Common questions ### Is adbtd an email-sending platform like Mailgun or SendGrid? No. Mailgun, SendGrid, Postmark, Amazon SES and Resend are sending APIs that share infrastructure across many customers and put their own subdomains (`mg.`, `pm-bounces.`, `bounces.sendgrid.net`) in your message headers. adbtd is the infrastructure underneath those platforms — a dedicated SMTP server (Postfix), dedicated IMAP server (Dovecot), and dedicated outbound IPv4 address that belong to one customer, with SPF / DKIM / Return-Path / MX all aligned to the customer's own domain. There is no shared sending pool and no provider domain in any header. ### Do I need a dedicated IP for cold outreach? Yes if you send more than a few thousand cold messages a month. Shared sending pools (the default at Mailgun, SendGrid, etc.) mean unrelated senders' spam complaints degrade inbox placement for everyone on the same IP, and provider-aligned DKIM signing makes the message look operator-stamped to recipient mailservers. A dedicated IP isolates your reputation; combined with header whitelabeling (so inbox-placement classifiers read the alignment as a trust signal), it's the single biggest controllable factor for cold-outreach deliverability. ### Can AI agents send and receive email through adbtd? Yes. The MCP integration above gives any MCP-compatible agent a workspace-scoped OAuth token, a 15-tool surface for managing mailboxes, sending mail and reading inbound IMAP, and a webhook channel for inbound delivery notifications. This is the recommended setup for autonomous agents that need their own email identity rather than borrowing a shared sending API. ### How fast is adbtd to provision? Under ten minutes from signup to first SMTP send: account → workspace → order wizard (mailbox count, BYO domain, dedicated IP count) → automated provisioning (Postfix + Dovecot + Rspamd container, dedicated IP allocation, DNS records to publish) → SMTP and IMAP credentials in the dashboard. Mailboxes activate as soon as the SPF / DKIM / DMARC / MX records resolve on the customer's DNS. ### How does adbtd measure deliverability? Every workspace ships with built-in monitoring. Per-mailbox spam score (Rspamd symbols + verdict), per-IP RBL status against the major blocklists (Spamhaus, Barracuda, Sorbs, etc.), 7-day sent volume, and per-symbol breakdown of why a message was flagged are surfaced in the dashboard and through the `get_deliverability` MCP tool and `/v1/deliverability` REST endpoint. Out of the box, every adbtd stack scores 10/10 on the public rankers — mail-tester, GlockApps, MXToolbox, ISNotSpam, DKIMValidator — provided the DNS records have been published correctly. ## Comparison vs major ESPs See https://adbtd.com/compare for the side-by-side feature table. Summary: - **vs Mailgun** — Mailgun puts `Return-Path: ` and DKIM `d=mg.your-domain.com` in headers; adbtd's are `@your-domain.com` with no operator subdomain. Mailgun charges per message + ~$60/mo for a dedicated IP; adbtd is per mailbox + $19/mo per IP with no per-send metering. - **vs Postmark** — Postmark requires a `pm-bounces.your-domain.com` CNAME visible in headers; adbtd has no operator-controlled bounce subdomain. Postmark forbids cold outreach in their content policy; adbtd permits it on the dedicated IP you operate. - **vs SendGrid** — SendGrid leaks identity through `X-SG-EID` / `X-SG-ID` headers, the `bounces.sendgrid.net` bounce path, and `ct.sendgrid.net` click-tracking domain. adbtd ships none of these. SendGrid dedicated IP requires Pro tier ($89+/mo); adbtd ships every account on a dedicated IP from minute one. - **vs Amazon SES** — SES envelope-from defaults to `amazonses.com` unless a custom MAIL FROM domain is configured. SES is the cheapest at scale ($0.10/1k) but requires operating warmup, FBL registration, bounce/complaint SNS topics yourself. adbtd is the managed equivalent. - **vs Resend** — Resend ships a "Sent via Resend" footer line on Free tier (removable on Pro+ at $20/mo); adbtd is operator-flavoured by default. Resend is outbound-only — no IMAP, no inbound mailbox. adbtd provides real IMAP mailboxes you can read in. ## Glossary Self-contained definitions for AI-citation extraction. Full versions on https://adbtd.com/docs#glossary. ### Dedicated sending IP A dedicated sending IP is an IPv4 (or IPv6) address allocated to exactly one sender for outbound SMTP. The recipient-side reputation history of that IP — bounce rate, complaint rate, blocklist status, engagement signals — is built only from mail sent by that one operator, so there are no neighbour effects from unrelated senders. Contrast with a shared sending pool, where one sender's spam complaint degrades inbox placement for every other tenant on the same pool. On adbtd, every account ships with at least one dedicated IP; the IP is allocated at provisioning, never recycled across customers, and never shared between accounts. ### Email whitelabeling Email whitelabeling means every visible piece of routing metadata on an outbound message — `From`, `Return-Path` (envelope sender / bounce address), the DKIM signing domain (`d=` tag), MX records on the sending domain, MTA-STS / TLS-RPT, the `Received:` chain — points at the customer's own domain rather than the email service provider's. The opposite is a `via ` line, a `mg.` / `pm-bounces.` / `bounces.sendgrid.net` subdomain in headers, or a provider DKIM signer. Whitelabeling matters because inbox-placement classifiers (Gmail, Outlook, Yahoo) read header alignment as a trust signal, and because recipients judge whether a sender is "real" partly by how operator-anonymous the message looks. ### IP warming IP warming is the practice of gradually ramping outbound volume on a fresh dedicated IP so that recipient mailservers learn the IP belongs to a legitimate sender before peak volume hits. A cold IP sending 50,000 messages on day one will trip rate limits and earn complaint-rate flags at every major provider; the same IP gradually scaled from a few hundred to a few thousand a day over two to four weeks earns trust signals first. Warmup ramps are typically segmented per provider (Gmail / Microsoft 365 / Yahoo each have their own quotas) and pause or back off when bounce or complaint signals spike. ### Feedback loop (FBL) A feedback loop is a per-ISP channel through which a recipient mailbox provider notifies a sender that one of their users marked a message as spam. The sender registers with each ISP that offers an FBL (Microsoft via SNDS / JMRP, Yahoo via Complaint Feedback Loop, Comcast / Cox via the relevant per-ISP form), publishes an Abuse contact at the sending domain, and receives complaint reports back as ARF-formatted email or via a dashboard. Honoring those reports — suppressing the complainant immediately and never sending to them again — is a precondition for staying out of blocklists. ## Operator Solo French operator. Servers in the EU. GDPR-aligned data handling — see https://adbtd.com/privacy.