Oracle Intelligencevia Pyth Hermes

CI Calibration Analysis

Measures how accurately Pyth's confidence intervals predict actual price moves. A perfectly calibrated oracle's ±1σ band should capture 68.3% of future prices — deviations reveal systematic over- or under-confidence.

Pyth Hermes API →

Ready to analyse

Select an asset, time period, and horizon above, then click Run Analysis.

Fetches historical snapshots from hermes.pyth.network and tests whether Pyth's published ±1σ CI captures 68.3% of real price moves — the first public empirical backtest of the Pyth oracle.