SEPE outages are now 503, not 502

19/09/2026

When SEPE goes down it does not always let us read its maintenance notice: sometimes its server stops answering verifiably and the connection is never established. Until today that came back like this:

Before

{
  "success": false,
  "message": "Error de conexión de SALTRA con el (SEPE). ServiceOffLine",
  "status": 502,
  "maintenance": false
}

Two problems for an integrator: the message names Saltra, as if the failure were ours, and maintenance is false, so there is no way to tell it apart from a one-off rejection that is worth retrying.

From now on it is an outage on their side like any other:

Now

{
  "success": false,
  "message": "El (SEPE) está en mantenimiento. Inténtelo más tarde.",
  "status": 503,
  "maintenance": true
}

It affects every SEPE operation, both queries and submissions.

What to check in your integration: if you decide whether to retry by looking at the message text, stop doing that and look at maintenance instead. That is the field meant for this, its wording does not change, and it already told apart every other outage on their side. The full table is in SEPE errors.