Transparent AI usage is a product feature.

Tracked Per AI Call

• Authenticated user id.
• App slug, provider label, model label, and request status.
• Input and output character counts.
• Timestamp and monthly usage bucket.

Saved App Results (Opt-in)

• When you run a built-in app with “Save to my history” checked, the input text and generated result are stored privately against your account.
• Unchecking the box stores nothing beyond the metadata above.
• History is capped at 50 entries per app and deletable at any time from the app’s History tab.
• Saved content is never shared with other users, and Krenos never uses it to train AI models.

Where AI Requests Go

Each request is processed by the AI provider that serves the model in use, under that provider’s terms. The sub-processors page lists every provider, where it processes data and whether it may train on inputs. Providers that process data in China or train by default stay off unless you turn them on in Account → Privacy.

Why It Matters

This gives users clearer cost and usage visibility than opaque AI limits. It also gives the platform a foundation for quotas, billing, abuse detection, and model-provider audits.