Fixed screening response schema, not arbitrary JSON Schema generation.
Generation controls
Seven current categories: crime, hate, profanity, sexual content, violence, weapons and self harm. Configurable L1–L4 sensitivity; higher levels include milder mentions. In Detect mode the top-level flagged field is always false even when detectors find a threat. Use breakdown results and the action field; Enforce mode permits flagged=true. Threshold labels are not calibrated numerical probabilities.Source for Generation controls (opens in a new tab)
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Processing region and log storage are independent. Log storage defaults to EU and is fixed when the organization is created. The PII guide says detected PII is masked before logging and models are not trained on PII; this does not establish zero retention or a universal no-training promise for all content. Exact retention and enterprise contractual terms remain unverified.
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Processing region and log storage are independent. Log storage defaults to EU and is fixed when the organization is created. The PII guide says detected PII is masked before logging and models are not trained on PII; this does not establish zero retention or a universal no-training promise for all content. Exact retention and enterprise contractual terms remain unverified.
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Processing region and log storage are independent. Log storage defaults to EU and is fixed when the organization is created. The PII guide says detected PII is masked before logging and models are not trained on PII; this does not establish zero retention or a universal no-training promise for all content. Exact retention and enterprise contractual terms remain unverified.
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Processing region and log storage are independent. Log storage defaults to EU and is fixed when the organization is created. The PII guide says detected PII is masked before logging and models are not trained on PII; this does not establish zero retention or a universal no-training promise for all content. Exact retention and enterprise contractual terms remain unverified.
No request model selector established. Policies select detectors; dev_info.model_version is response build metadata.
Access
Enterprise license and private registry credentials required. Helm, Docker and air-gapped deployments; exact container revision not established.
Regions
Customer infrastructure
Endpoint pricing has not been collected. Check the provider for applicable rates.
Data & enterprise terms
Processing region and log storage are independent. Log storage defaults to EU and is fixed when the organization is created. The PII guide says detected PII is masked before logging and models are not trained on PII; this does not establish zero retention or a universal no-training promise for all content. Exact retention and enterprise contractual terms remain unverified.
Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.
01 / INITIAL INPUTProcess & retain
The system processes eligible input and retains reusable state.
02 / LATER REQUESTCheck available state
Reuse depends on matching input, session state and the cache policy.
03 / CACHE HITReuse & generate
Reusing state can reduce repeated work. Any charges follow the billing rules below.
AI Guardrails API (global)
Not verified
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How it works
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Minimum reusable input
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Lifetime & refresh
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How to enable it
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Usage fields
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Billing rules
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Limits & data boundaries
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
AI Guardrails API (eu)
Not verified
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How it works
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Minimum reusable input
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Lifetime & refresh
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How to enable it
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Usage fields
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Billing rules
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Limits & data boundaries
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
AI Guardrails API (us)
Not verified
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How it works
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Minimum reusable input
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Lifetime & refresh
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How to enable it
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Usage fields
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Billing rules
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Limits & data boundaries
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
AI Guardrails API (singapore)
Not verified
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How it works
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Minimum reusable input
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Lifetime & refresh
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How to enable it
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Usage fields
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Billing rules
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Limits & data boundaries
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
AI Guardrails self-hosted
Not verified
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How it works
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Minimum reusable input
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Lifetime & refresh
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
How to enable it
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Usage fields
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Billing rules
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
Limits & data boundaries
Not collected
No documented model-specific prompt-cache mechanism or tariff established. Client HTTP caching is not inference caching.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 15 Oct 2026
What the measurements say
No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.
Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.
The record behind the model
Why it is included
Provider documentation explicitly describes trained security model families. Deployment requires an Enterprise license and provider-issued container registry credentials. This record represents the private model family, not a publicly identified checkpoint or an API version.
