Cerbos

Founded 2021

Manage fine-grained access control and evaluate authorization decisions for and AI agents

Pricing model

Free Tier, Paid

You can use Cerbos to manage fine-grained access control across your applications, services, and AI agents. It lets you write authorization policies in human-readable YAML files, test them centrally, and enforce them at runtime using stateless decision points. The system evaluates roles, attributes, and relationships to deliver access decisions in under a millisecond. It separates access rules from application code, lets you update permissions without redeploying software, and generates detailed audit logs for every allow or deny decision to ensure compliance.

What Cerbos does

Define role-based and attribute-based permissions in YAML filesQuery access decisions from backend code using client SDKsInspect audit logs for specific user and machine access attemptsSimulate and test policy adjustments before deploying to productionSet runtime permission boundaries for autonomous AI agentsDeploy stateless policy decision points into Kubernetes clustersHuman-readable YAML policy definitionsSupport for RBAC, ABAC, ReBAC, and PBAC in a single languageSub-millisecond access evaluation via stateless decision enginesCentral policy authoring, testing, and versioning hubAI agent and Model Context Protocol guardrail managementGranular audit logs for all allow and deny actionsSDKs available for major programming languagesFlexible deployment on cloud, on-premises, or air-gapped systems