Meta Llama API

Where does this provider put your AI data — and who can compel its disclosure?

ResidencyUnited States us
SovereigntyUnited States us
Endpointsapi.llama.com

Data practices

How does this provider treat the data you send it? Advisory statements about documented practices — not legal advice.

Trains on customer API data by defaultNo source — Model API Terms of Service section 5.1: 'Meta will not use Content from Standard Services to train Meta Models' (updated 2026-08-21) (retrieved 2026-08-23)
Retention (API inputs/outputs)Content and Usage Data retained as needed to provide the Services, for legal compliance, in connection with Meta's rights under the terms, when systems flag data for a potential policy violation, and for security/safety/abuse/policy review. No fixed retention window is published. source — Section 4.4 (Retention): five enumerated retention bases; no fixed window published (retrieved 2026-08-23)
Subprocessor listhttps://forwork.meta.com/legal/mfw-subprocessor-list/ — Meta's published subprocessor list, referenced from Model API Terms of Service section 5.1 (Standard Services processor terms); may be updated by Meta from time to time (retrieved 2026-08-23)
Enterprise tierTier split: Standard Services — 'Meta will not use Content from Standard Services to train Meta Models' (section 5.1). Discounted Services (discounted model tier) — content MAY be used to train, develop and improve Meta AI models by default, disassociated from your account/key before training use (section 6.1); traffic requiring no-training must use Standard Services. source — Section 6.1 (Discounted Services): 'You agree that Meta may use your Content to train, develop, evaluate, and improve Meta's artificial intelligence models, products, and services', disassociation steps before training use; section 6.2 directs confidential traffic to Standard Services (retrieved 2026-08-23)
Entry last reviewed2026-08-23