Backfrom OpenRouter’s own policies · verified 2026-09-02

OpenRouter: data residency & retention

Headquarters
🇺🇸 United States (US)
Where inference runs
Not disclosed
OpenRouter runs no inference itself — inputs are transmitted to the model provider you select, and each downstream provider's retention/training policy is surfaced per endpoint.
Region pinning
Yes, self-serve
Self-serve routing constraints via provider allow-lists, 'data_collection: deny', and the 'zdr' parameter; stricter EU in-region routing through eu.openrouter.ai is enterprise-only.
Prompt retention
Not retained
OpenRouter itself has a ZDR policy — prompts are not retained unless you opt in to prompt logging; retention at the model provider is governed by that provider's own policy, shown per endpoint.
Trains on API data
No
OpenRouter does not use inputs or outputs for model training; opting out of training in account settings stops routing to providers that train.
Zero data retention
Available
Account-level ZDR setting plus a per-request 'zdr: true' parameter restrict routing to endpoints with zero-data-retention policies.

Compiled from OpenRouter’s published privacy policy, terms, and documentation on 2026-09-02. “Not disclosed” means OpenRouter publishes nothing on the point. Policies change without notice, so verify against the sources above before relying on this for compliance decisions.

Standard Compute

Plug in Standard Compute wherever you need intelligence.

One key. Every frontier model. A bill that stays put.

7-day fair refund Cancel anytime Billing by Stripe