Poste
Email addresses for AI agents.
An agent can create and pay for its own mailbox using USDC over x402. No account or API key is needed.
agent poste sender
│ POST /v1/mailbox │ │
│ ─────────────────────────▶ │ │
│ 402 · pay $0.05 (x402) │ │
│ ◀───────────────────────── │ │
│ retry + payment │ │
│ ─────────────────────────▶ │ address + token │
│ ◀───────────────────────── │ │
│ ( the agent signs up; the sender emails it ) │
│ │ ◀───────────────────── │
│ │ parse · extract code │
│ GET /v1/mailbox/code │ │
│ ─────────────────────────▶ │ │
│ { code: "482913" } │ │
│ ◀───────────────────────── │ │
Get an address
POST https://api.poste.sh/v1/mailbox
pay the 402 with any x402 client
→ { address, token }
GET https://api.poste.sh/v1/mailbox/code
→ { code }
POST https://api.poste.sh/v1/mailbox/send
Incoming email is parsed automatically. If it contains a verification code, Poste extracts it so the agent can get the code in one call.
Errors also explain what to do next.
Full reference: llms.txt
No wallet yet?
Use:
POST https://api.poste.sh/v1/mailbox/reserve
This returns the address and token immediately, together with a payment link.
A person can complete the payment in a browser wallet. Incoming mail is held until the mailbox is paid.
Prices
burner 1 hour, receive only $0.05 identity 30 days, send + receive, 200 sends $1.00 renew 30 days $0.50 send per message after 200 $0.005 read always free
Prices are per mailbox and paid in USDC on Base Sepolia testnet.
Built for real use
Mailboxes are paid, which makes abuse more expensive and helps the addresses work with signup forms.
Sending is rate-limited. The From address cannot be spoofed. Complaints can suspend sending.
Incoming mail also includes its SPF, DKIM and DMARC results.
Domains are published and abuse@poste.sh is monitored.