Skip to main content
POST
Retry Delivery
Re-send a delivery that failed, or one that succeeded but was lost downstream. The payload and its event_id are unchanged, so a handler that dedupes on the event id can safely receive a replay. Replaying costs no credits — you never need to re-run a job to recover a missed event. Retrying resets the attempt counter, so the delivery gets the full 30s / 120s / 480s ladder again.

Path parameters

string
required
Delivery UUID from List Deliveries.

Response (200 OK)

The delivery, now pending with attempts reset to 0. Same shape as List Deliveries.

Errors

error
No such delivery, or it belongs to another API key.
error
Already queued for another automatic attempt. Wait for it to resolve — next_retry_at tells you when.
error
The endpoint has been deleted or deactivated, so a replay would fail for the same reason. Register a new endpoint first.

Recovering after an outage

Deploy the fix before replaying in bulk. A replay against a still-broken endpoint burns the retry ladder again and lands back in failed.