Where does this provider put your AI data — and who can compel its disclosure?
| Residency | United States us |
|---|---|
| Sovereignty | United States us |
| Endpoints | api.deepinfra.com |
| Packages | @ai-sdk/deepinfra (npm) |
How does this provider treat the data you send it? Advisory statements about documented practices — not legal advice.
| Trains on customer API data by default | No source — Terms of Service (last modified 2026-08-17) §7(b): 'Provider will not sell Customer Data, and will not use Customer Data to train, fine-tune, or otherwise improve any model, except as necessary to provide the Services' (retrieved 2026-10-10) |
|---|---|
| Retention (API inputs/outputs) | Contractual zero data retention: customer data submitted to or generated by the services is not retained, stored, or logged 'beyond the period strictly necessary to process and return the applicable request' (ToS §7(b)). Exceptions: data retained at the customer's written request for support is deleted within 30 days of resolution; image-generation outputs are stored 'for a short period'; bulk-inference payloads may be stored encrypted on disk for an unspecified short period; non-content metadata (request ID, cost, sampling parameters) is logged. source — §7(b): 'Provider will not retain, store, or log any Customer Data… beyond the period strictly necessary to process and return the applicable request, after which such Customer Data is deleted… ("Zero Data Retention")', with the support-request (30-day), legal, and fraud/security exceptions (retrieved 2026-10-10) |
| Subprocessor list | https://docs.deepinfra.com/account/subprocessors — Published list (last updated 2024-09-06): Stripe (payments), Amazon Web Services and Google Cloud Platform (infrastructure) (retrieved 2026-10-10) |
| Enterprise tier | No tier changes the defaults — no-training and zero retention already apply at the base tier; an executed DPA 'supplements, and does not narrow' the ZDR commitment and controls over the Privacy Policy for all customer data (ToS §5(a)). Enterprise options are dedicated/private deployments ('data stays on dedicated infrastructure') and DeepCluster dedicated GPU clusters. Partner-model caveat: requests routed to Google or Anthropic models follow those companies' training policies, not DeepInfra's. source — 'No training': 'We do not use data you submit to our APIs for training models, except when using Google or Anthropic models, where the receiving company's training policy applies' (retrieved 2026-10-10) |
| Entry last reviewed | 2026-10-10 |