Skip to content
aityx
Esc
↑↓navigate↵open⌘Jpreview
On this page

aityx API

System One is the only public operation. It executes a supplied decision with matching facts, forwards Jev questions, or feeds Jev answers into a supplied decision. All three shapes use $0.042 per million API input tokens with output free. There is no model-authoring or input-interpretation endpoint; to run unstructured input through a decision, send Jev questions with it. Public execution returns answers, receipt and usage only, plus decision_error on a native failure after Jev completes. Timing, provider counts, charges and balances are console/internal details. Jev usage uses exact provider counts; combined calls add native tokens without counting shared data twice. Native definitions are supplied in decision with one state; model is reserved for a Jev model alias; answer types are inferred from named decisions. Execution returns a trace without echoing models or inputs. Your application keeps definitions, original state and receipts; there is no durable model or receipt storage. Model checks, replay, accounts and balance management remain console/internal operations.

Version 0.98
Base URLhttps://api.aityx.ai

Operations

Was this page helpful?