Mercury API

Email verification for developers

Mercury is a hosted email verification API. One request tells you whether an address is likely to reach an inbox — before you send.

curl "https://mercury-api.deepeshkalurs.workers.dev/v1/[email protected]" \
  -H "Authorization: Bearer mr_live_xxxxxxxxxxxxxxxx"

{
  "email": "[email protected]",
  "verdict": "valid",
  "risk_score": 7,
  "checks": {
    "format": { "ok": true },
    "mx": { "ok": true, "hosts": ["alt1.gmail-smtp-in.l.google.com"] },
    "disposable": { "is_disposable": false, "source": "vendored+custom" },
    "role": { "is_role": false, "role": null },
    "smtp": { "ok": true, "code": 250, "latency_ms": 312 },
    "reputation": { "score": 12, "risk": "low", "signals": [] }
  },
  "cached": false,
  "cache_ttl_seconds": 3600,
  "verified_at": "2026-08-02T10:21:00Z"
}

Quickstart

Verify your first email in under a minute — sign up, create a key, call the API.

Read the quickstart →

API reference

Every endpoint, parameter and response, generated from the OpenAPI document.

Browse the reference →

The pipeline

format → disposable → role → mx → catch-all → reputation → SMTP probe. Cheap checks first, short-circuit on hard fails, verdicts are never binary.

How verification works →

Why teams use Mercury

Verdicts, never binary. valid, invalid, risky, unknown — plus a 0–100 risk_score so your integration decides what to do, not us.
Sub-200 ms p50. KV-cached results with parallel MX via DoH; cache hits are free.
Built for deliverability. Disposable + role detection and Spamhaus-based reputation signals protect your sender score — not just your sign-up form.
An API first. OpenAPI 3.0 document, typed SDKs, webhooks, and a tiered pricing model that starts at $0 forever.

1,000 free lookups / day

The Free tier is the trial — it never expires and never asks for a card.

Create a team →

Bulk verification

Verify up to 100 addresses per call (Pro) or 1,000 (Team) with a single async job and webhook delivery.

Bulk endpoint →

Status & transparency

Same SLA language on every tier, public status page, and a SOC 2 Type II audit in progress for Q4 2026.

Browse the reference →