AirLLM security & data privacy Enterprise-Grade Security
AirLLM never uses any prompt to train models, on any plan, and on paid plans it avoids model providers that use prompts to train models. Security and privacy are designed in from the ground up — not added on.
0
AirLLM training use
RBAC
Role-based access
TLS
In transit
SSO
Sign-in
Security Features
Reasonable, layered safeguards across the platform.
Data Encryption
Encryption in transit with TLS, and reasonable safeguards for stored data. Additional key-management options can be discussed as part of an enterprise agreement.
Limited Prompt Storage
Prompts and responses sent through Air Router or the AirLLM API pass through short-lived, account-scoped caches that expire automatically — generally within 24 hours of last use. When a request fails, AirLLM may keep its content, with secrets removed, to diagnose the failure, and it keeps samples of responses to check routing quality; these records are deleted with your account's data. A copy attached to an engineering ticket is kept only as long as needed to fix the problem. We also keep usage metadata, such as token counts and cost, and support chats with Alina. Our commitment: AirLLM never uses any prompt to train models, on any plan, and never sells your content. On paid plans, AirLLM avoids model providers that use prompts to train models; on the Free plan, model providers may use your prompts to train models and may retain them. See our Data Retention Policy.
SSO & RBAC
SAML 2.0 and OIDC single sign-on with granular role-based access control. Define who can access which models, features, and data within your organization.
Network Security
Encrypted connections to the gateway, with enterprise network options such as IP allowlisting and private connectivity available to discuss as part of an enterprise agreement.
Audit Logging
Usage records for API calls and account activity, visible in your dashboard. Enterprise log-export options can be discussed as part of an enterprise agreement. On our side, administrative access to the AirLLM control plane requires a named operator account with multi-factor authentication; access follows least privilege; administrative actions are logged with the operator's name.
Incident Response
Automated monitoring, and we aim to respond to security incidents promptly. If a security incident affects your data, we notify you without undue delay, as required by applicable law.
How Your Data Stays Private
A transparent look at our data handling architecture.
Your Application
Sends API request
AirLLM Gateway
Routes it, caching briefly
Model Inference
Under the provider's terms
Questions About Security?
Our security team is available to answer your questions, provide documentation, and support your vendor security review process.
Contact Security Team