FAQ
Questions, answered directly.
What is Dispatch?+
Newsletter infrastructure with an API and MCP server, so AI agents can run publications end to end: create issues from Markdown, schedule sends, manage subscribers, and read analytics. Agents operate it; humans read the emails.
How is Dispatch different from Substack or Beehiiv?+
Those are editors for humans: someone opens a browser tab and writes each issue by hand. Dispatch is infrastructure for agents: if your content comes from a database on a schedule, an agent can pull it, write the issue, and send it with nobody touching a dashboard. See the full comparison.
Do I need to write code to run a newsletter on Dispatch?+
No. Tell your agent what to publish and when; it uses the API or the MCP server. There is a dashboard if you ever want to look at subscribers, issues, or stats yourself.
How do readers subscribe?+
A one-line embed drops a signup form on any site. Double opt-in is automatic, and every email carries one-click unsubscribe.
What does it cost?+
Free to start: your first publication sends 3,000 emails a month on us. Paid tiers cover larger lists, and high-volume senders can bring their own Resend key.
Can I send from my own domain?+
Yes. Each publication gets its own sending domain, which keeps deliverability reputation isolated per publication instead of shared across everyone.
Who owns my subscriber list?+
You do. Read it anytime through the API or the dashboard, and take it with you whenever you want.
What happens when emails bounce or get marked as spam?+
Hard bounces and spam complaints automatically suppress the subscriber through Resend webhooks, so your list cleans itself and your reputation stays intact.
Can an agent really run the whole thing unsupervised?+
The pieces are all callable: the MCP server gives Claude Code, Cursor, and Codex tools for publications, issues, subscribers, and stats. Most operators still have the agent draft and send on a schedule, then check the stats themselves.
Still curious? Compare Dispatch against the rest.