Data and security

Content guardrails

PII, secret, prompt-injection and safety checks you configure per identity on requests and completions.

Content guardrails inspect a chat request after authentication and before routing, and the completion before it is returned. You configure them per identity on the API keys page, Guardrails tab. They are separate from spend and rate limits, which live under Budget & limits.

Classifiers and policies

PieceWhat it is
ClassifierWhere a check runs: an HTTPS endpoint you operate (JSON in, verdict out), or a keyword list
PolicyAt most one per identity: the standard pack (seven ordered checks) or your own chain
CheckA stage (input or output), a capability, an action (block or redact) and a timeout
ProtectedOn a classifier timeout or failure: refuse the request, or skip the check

A policy on the default identity governs every key created without an explicit identity. A policy can be paused without deleting it.

At a glance
ConfiguredAPI keys > Guardrails
ScopePer identity
Default check timeout250 ms
SurfacesChat Completions, Responses, Messages
Content storedNone, decision counts only

What a block looks like

An input violation never reaches a provider. An output block buffers the completion and refuses it. Redaction rewrites message or completion text; tool-call arguments are never rewritten, so a redaction on a completion with tool calls becomes a block.

OutcomeResponse
Policy violation400 content_filter
Classifier timeout or failure, protected policy503, retryable
Classifier failure, unprotected policyCheck skipped
{
  "error": {
    "type": "invalid_request_error",
    "code": "content_filter",
    "message": "..."
  }
}

Coverage and cost

SurfaceBehavior under an enforcing policy
Chat Completions, Responses (HTTP and WebSocket), MessagesInspected
Embeddings, images, audio, batchesRefused, because they cannot be inspected
Observation-only policyRecords what would happen, never refuses

Protected checks add latency before the first byte: up to the sum of the input checks' timeouts. Prompts, completions and detector payloads are never stored; the tab shows a seven-day count of blocked, redacted and failed checks per identity.

Your classifier endpoint
AddressPublic HTTPS only
CredentialBearer token, stored encrypted, last four shown
Redirects, private IPsRefused

Not the same as content moderation

Uncensored models carry a platform moderation check that you do not configure. Guardrails are yours: you choose the checks, the classifiers and the identities they cover.

GuardrailsModeration
Who configuresYouExperiential
Applies toYour identitiesUncensored models
Stores contentNoNo