Developers
Verify a user in two API calls.
REST over HTTPS, JSON in and out, bearer keys scoped per app and environment (mv_test_ / mv_live_).
Endpoints
- POST /v1/verify
Start a verification. Runs fraud pre-flight, then cascades through channels.
- POST /v1/verify/{id}/check
Submit a code. Single-use, max 5 attempts, constant-time compare.
- POST /v1/verify/{id}/resend
Resend on the next channel. Bounded by cooldown and resend limit.
- GET /v1/verify/{id}
Fetch status, channel path, latency and cost. Never returns the code.
- POST /v1/verify/{id}/cancel
Cancel a pending verification.
- GET /v1/signals
List signal logs with filters for channel, status, country and fraud score.
- POST /v1/webhooks/{event_id}/replay
Replay a webhook delivery to your endpoint.
- GET /v1/wallet
Balance, spend by channel and SLA refund credits.
Webhook events
Signed with HMAC-SHA256 in the MoVerify-Signature header. Retried with backoff for 24h.
verification.startedverification.channel_failedverification.deliveredverification.verifiedverification.expiredverification.blockedwallet.low_balancekey.circuit_breaker_tripped
curl -X POST https://moverify.vercel.app/api/v1/verify \
-H "Authorization: Bearer mov_test_••••" \
-H "Content-Type: application/json" \
-d '{
"to": "+27782810441",
"country": "ZA",
"channels": ["whatsapp", "sms"]
}'
# 2. Check submitted code:
curl -X POST https://moverify.vercel.app/api/v1/verify/check \
-H "Content-Type: application/json" \
-d '{"id": "vr_••••", "code": "849201"}'POST /v1/verify parameters
| to | string | E.164 phone number, or email when channel is email |
| channels | string[] | Ordered list: whatsapp, push, voice, sms, email. Default: smart |
| code_length | 4–8 | Default 6 |
| expires_in | seconds | 60–600. Default 300 |
| timeouts | object | Per-channel failover timeout, e.g. { whatsapp: 8, sms: 15 } |
| locale | string | en, fr, ar, sw, pt, ha, yo, ig, zu |
| metadata | object | Your reference data, echoed in webhooks |
Security defaults
- Codes from a CSPRNG, stored only as salted hashes
- Single-use, expire in 5 minutes, 5 attempts max
- Per-key, per-number and per-IP rate limits
- Plaintext codes and secrets never written to logs
- Keys isolated per app — one product cannot read another's data
