Backfrom NVIDIA’s own policies · verified 2026-09-02
NVIDIA: data residency & retention
Headquarters
🇺🇸 United States (US)
Where inference runs
Not disclosed
The API Trial Terms do not state where api.nvidia.com inference runs. NVIDIA's privacy policy says only that 'in most cases we need to securely transfer and store your information in the United States' (about personal information generally, not a per-request inference-region commitment).
Region pinning
No
No region selection or residency controls exist for the trial API catalog.
Prompt retention
Not retained
API Trial ToS §2.3: 'NVIDIA will not store or use User Content or Generated Content at the end of each API Service session', except (§2.4) fine-tuning-type services (User Content 30 days, Generated Content 90 days) and security/fraud/abuse logging (§3.3).
Trains on API data
Yes, by default
API Trial ToS §3.3(iv): NVIDIA collects, 'without identifying specific users... User Content and Generated Content to improve NVIDIA products and services, including AI models.' This sits alongside the §2.3 no-storage default; production use under a paid Subscription is governed by separate terms.
Zero data retention
Not offered
No formal ZDR program; stateless no-storage after each session is the stated default of the trial terms, but §3.3 improvement/training collection and security logging still apply.
Compiled from NVIDIA’s published privacy policy, terms, and documentation on 2026-09-02. “Not disclosed” means NVIDIA publishes nothing on the point. Policies change without notice, so verify against the sources above before relying on this for compliance decisions.