What is an agent inbox?
An agent inbox is an email address that belongs to an AI agent or an app, not to a person. Here is what that means in practice, how it differs from a normal mailbox, and when you need one.
By Arin Prajapati, founder, 3 min read
On this page (5)
The short definition
An agent inbox is an email address that belongs to an AI agent or an app rather than to a person. You create it with an API call, the agent reads and sends mail through that API or through a tool interface such as MCP, and the inbox can be thrown away when the task ends. It exists because agents keep running into the same wall: the world signs you up by email, and an agent has no mailbox.
How it differs from Gmail
A personal mailbox assumes a human is looking at it. That shows up as a sign in flow, a consent screen, rate limits tuned for people, and mail you have to read as text. An agent inbox changes each of those:
- Created in one call, with no sign in and no consent screen, so an agent can make a new address for each task.
- Read as data. A message arrives with its sender, thread, and any one time code or confirmation link already pulled out.
- Waited on, not polled. The agent makes one blocking call that returns when the next matching mail arrives, which saves steps and tokens.
- Scoped. A key can reach one inbox and nothing else, so a mistake cannot read the rest of your mail.
What an agent inbox needs
If you are evaluating one, check for an inbox you can create on demand, receiving that works on your own domain, a blocking wait with extracted codes, a tool interface for your agent client, scoped and revocable keys, and cheap addresses. We list these as a checklist in the best email API for AI agents, and explain why each one matters in giving AI agents an email address.
Runearc is currently free. The paid Runearc prices below are planned and come from our pricing page; billing is not live yet.
When you need one
- An agent signs up for a service and must read the verification email.
- You run end to end tests that need a fresh address and the link from the email.
- An agent should send and reply from its own address, for example a support or scheduling assistant.
- You build a product where each customer or each agent run needs its own address.
You do not need one to send app email such as receipts, and an agent inbox is not a way to read someone else's existing Gmail or Outlook mailbox; that is a different kind of API. To try it, follow the walkthrough or connect through MCP. Runearc is agent inboxes, sending and a support inbox in one account, with a free plan.
Questions
What is an agent inbox?
An agent inbox is an email address that an AI agent or an app owns, created and read through an API instead of a mail client. It receives mail, can send and reply, and lets the agent wait for the next message, such as a verification code, without polling.
Is an agent inbox the same as a Gmail account?
No. A Gmail account is built for a person: sign in screens, consent prompts and limits designed for humans. An agent inbox is created in one API call, scoped to one agent, and read through structured fields such as extracted codes and links.
Do I need an agent inbox to let an agent sign up for a service?
If the service sends a verification email, yes: the agent needs an address that receives mail and a way to read the code. A catch-all domain lets the agent invent a fresh address per task.
Can I use my own domain for agent inboxes?
Yes. Agent mail from your own verified domain looks more trustworthy than mail from a shared domain. Runearc also gives every workspace a private runearc.email domain that works immediately.
Arin Prajapati, founder of RunearcA software engineer in Mumbai who builds Runearc's mail servers, API and site. More about Arin and why he built Runearc.
Read next
- The best email API for AI agents: what to look forAn agent needs to sign up, read a code and reply, without a human glancing at a mail client. These are the seven things to check in an email API for agents, and how the options measure up.
- An email inbox for your AI agentAn agent that signs up for things needs an address that receives mail, and a way to get the code out of it without polling. Here is the whole flow in three steps.
- Give an MCP agent its own email addressMCP lets an agent use email the way it uses any other tool. Add one server to your client and the agent can create an inbox, send, reply and wait for a code.
Put Runearc to work
Get inboxes, sending, your own domains and MCP in one workspace. The free plan includes 3 inboxes, 3,000 emails a month and 3 custom domains, with no card.