Updated June 17, 2026
AI Processors
nagu uses AI processors only to power user-facing inbox features for the connected account. This page is generated from the same runtime manifest that routes live AI traffic, so the list below matches the processors actually in use.
Zero Day Data Retention for LLMs
For LLM processing, nagu uses a Zero Day Data Retention posture: prompts, message excerpts, tool context, and model outputs are processed to return the requested feature and are not retained by the provider.
Three separate promises about training
- No general model training. We do not use your mail content to train or improve general-purpose AI models, ours or anyone else's.
- Per-user personalization only. nagu learns your voice and habits to power your own features — a per-user, account-scoped model that is yours alone, not a shared one, and you can clear it at any time.
- Provider no-training, by contract. Every AI provider we route to is contractually blocked from training on your content; LLM traffic carries a no-training, zero-retention posture.
These are three different statements. The first is about general models; the second is about a feature that belongs to you; the third is a contractual posture on the providers below. None of them imply the others.
Current AI processors
This table is generated from nagu's runtime compliance manifest — the same source that decides where AI requests are routed — so it stays in sync with the providers actually in use. Every route below carries a no-training, zero-retention posture.
| Processor | Role | Region | Retention and training |
|---|---|---|---|
| Vercel AI Gateway | Model routing only — the gateway in front of every provider. Not a model-training destination for your content. | United States | Zero-day retention — content is processed for the request, then dropped. No training on your content. |
| Anthropic | Language-model processing for triage, briefs, summaries, and reply drafts when routed through the gateway. | United States | Zero-day retention — content is processed for the request, then dropped. No training on your content. |
| OpenAI | Language-model processing for triage, briefs, summaries, and reply drafts when routed through the gateway. | United States | Zero-day retention — content is processed for the request, then dropped. No training on your content. |
Data minimization
nagu sends only the account-scoped excerpts, metadata, instructions, and context needed for the requested AI feature. We do not send an entire mailbox when a single thread, summary, draft, or brief can be produced from a narrower context.
Changes
Because this page reads from the runtime manifest, the list updates when the providers in use change. Material changes to AI processing are also reflected in the Privacy Policy and the how your mail data moves overview.