Openlayer Guardrails

Openlayer Guardrails inspects live AI inputs and outputs and can block, redact, escalate, or route policy violations for review. It targets prompt injection, jailbreak attempts, exposure of personal or health information, and harmful or toxic content; Openlayer says it detects and blocks more than 50 PII and PHI entity types inline. Teams can configure which sensitive-data categories to block and topics an assistant should avoid. Guardrails are commonly enforced through the Gateway or applied directly through the SDK. Each decision can be logged with a timestamp, system version, owner, supporting trace, and reason for blocking. The product connects through SDKs, framework integrations, a network-level gateway, a CLI, and a REST API. Deployment options include a customer VPC and air-gapped environments; on-premise deployment is available on Enterprise. The free Basic plan is limited to one member, five projects, one inference pipeline per project, 20,000 inference logs per month, 20 tests per project, one workspace, and three months of data retention. Enterprise has custom pricing and includes unlimited members and projects, custom pipelines and logs, unlimited tests, and on-premise deployment.