Documentation

Trust Levels

How Cailos classifies and enforces data handling guarantees across providers.

Every endpoint in Cailos is assigned a trust level from 0 (lowest) to 3 (highest). This rating reflects the provider's data handling practices, security certifications, and deployment model. Trust levels are a hard routing constraint — when you request a minimum trust level, your data is never sent to endpoints below that threshold.

Levels

0 Low

No specific data handling guarantees from the provider. Data may be used for model training, stored indefinitely, or processed in jurisdictions without data protection laws. Suitable for non-sensitive, public data only.

Typical providers: community-hosted open-source models, free tiers with no SLA.

1 Standard

Data encrypted in transit (TLS). Provider operates under standard commercial terms of service. Data retention and training policies are documented but may allow provider use of inputs/outputs. Suitable for general business data.

Typical providers: major cloud providers with standard API terms (OpenAI default, Groq, Together, etc.).

2 High

SOC 2 Type II and/or GDPR compliant. Provider explicitly commits to not training on customer data. Data retention is limited and documented. Inputs/outputs are not stored beyond the request lifecycle unless explicitly opted in. Suitable for confidential business data, PII with consent, and regulated industries.

Typical providers: Anthropic (with data retention off), OpenAI Enterprise, Google Cloud AI with DPA.

3 Maximum

Zero-knowledge inference, hardware-attested secure enclaves, or on-premise deployment. The provider cannot access plaintext inputs or outputs, even in principle. Data never leaves a trusted execution environment. Suitable for classified data, sensitive defense applications, healthcare PHI, and financial data requiring maximum confidentiality.

Typical providers: Tinfoil (enclave-verified), self-hosted models, on-premise deployments.

How routing works

Trust is enforced as a hard filter in the routing pipeline. When you include trust_level in your request, Cailos eliminates all endpoints below that threshold before any ranking or scoring occurs. This means:

Requesting trust_level: 2 routes only to endpoints with trust 2 or 3.
If no endpoint meets the threshold, the request fails with a clear error rather than falling back to a less trusted provider.
Trust filtering happens before cost, speed, or quality optimisation — it is never overridden by routing preferences.

Usage

Pass the trust_level hint in the cailos object of your request:

{
  "model": "auto",
  "messages": [{"role": "user", "content": "..."}],
  "cailos": {
    "trust_level": 2,
    "optimise": "quality"
  }
}

Tier restrictions

Free-tier teams are restricted to trust level 0 regardless of the requested level. Paid teams can request up to trust level 3. This is enforced server-side and cannot be bypassed by the client.

Assignment criteria

Trust levels are assigned by Cailos administrators based on provider documentation, certifications, and deployment model. The assignment considers:

Data retention policy — Does the provider store inputs/outputs? For how long?
Training policy — Are customer inputs used to improve models?
Certifications — SOC 2, ISO 27001, GDPR compliance, FedRAMP, etc.
Deployment model — Shared cloud, dedicated instance, on-premise, or secure enclave.
Jurisdiction — Where is data processed and stored?

← Back to Cailos API Docs →