First-party data providers / publishers: Major exchanges, market makers, trading firms provide price data directly. This means Pyth tries to avoid too many intermediaries or opaque aggregations.
Aggregation algorithm: It combines multiple inputs, weighs them (depending on publisher reputation, stake etc.), filters out outliers, to produce a reliable price feed with a confidence interval.
Pull-oracle architecture: Instead of pushing price updates continuously (which can incur gas / compute / transaction costs even when nobody needs the update), Pyth lets data users request (pull) the latest data when they need it. This saves costs and improves efficiency.
Total Value Secured (TVS): At various points in growth: passed 1.5B across ~120 protocols in earlier phases. More recent reports show TVS recovering to several billions.
Transaction / Trading Volume Secured: In March 2024, Pyth secured $87.1B in volume (TTV) yearly, up massively from 4.8B the previous year.
Number of Price Feeds / Blockchain Integrations: 200+ price feeds; support for 40-50+ blockchains; hundreds of applications (250+ apps) using Pyth feeds.
