Send API
One endpoint for transactional messages, with API keys, delivery IDs, and logs.
Send and receive email, let AI draft replies, and control what gets sent—from one API and dashboard. Track messages, approvals, and decisions in one place.
POST /api/v1/emails
{
"from": "team@yourdomain.com",
"to": "customer@example.com",
"subject": "Welcome",
"text": "Your account is ready."
}Use the email API for your application’s messages. Add AI reply drafting when you want it, with rules that keep you in control.
Verify your sending domain to send from your own addresses. Configure receiving when you want incoming email to reach the platform.
Create an API key and send welcome emails, notifications, and other transactional messages. Track their activity in the dashboard.
Assign an AI agent an email identity and instructions for drafting replies to incoming messages. AI drafting requires a separately enabled allowance.
Set policies for the agent’s replies: allow sending, require human approval, or block sending. When no policy matches, human approval is required.
Review conversations, delivery activity, approvals, and the audit trail together. Follow a reply from its draft to its recorded outcome.
Email plans do not include AI drafting. New workspaces start with 0 AI drafts. See allowances and pricing →
Everything developers expect from email infrastructure, with controls for autonomous software.
One endpoint for transactional messages, with API keys, delivery IDs, and logs.
See verified sending domains and their receiving capabilities in one workspace.
Turn replies and new messages into structured events your application can process.
Register endpoints for received, sent, and failed email events.
Give your AI agent an email identity and let it draft replies to incoming messages. Your policies determine which replies may be sent, which need approval, and which are blocked. Follow every decision in the audit trail.
AI-generated replies can be wrong. Require human review when accuracy, tone, or sensitive information needs attention. Policies control whether a reply may be sent; they do not guarantee its content is correct.
Open the developer dashboard →