Pyth Network’s pull-based model (also called “pull oracle” or “on-demand” model) works as follows:

In contrast to traditional push-based oracles (like Chainlink’s primary model), where data providers continuously broadcast (“push”) price updates on-chain at fixed intervals regardless of demand — incurring high ongoing gas costs and scalability limits — Pyth uses a pull-based approach.

•  How it operates: Price data from first-party sources (exchanges, market makers like Jane Street/CBOE, DeFi protocols) is aggregated off-chain (primarily on Pythnet, a Solana-based appchain) in real-time with sub-second latency (~400ms updates). The on-chain price isn’t constantly updated; instead, it’s stored/updated only when requested (“pulled”).

•  Permissionless updates: Anyone (users, dApps, bots, or relayers) can permissionlessly submit the latest aggregated price update to the Pyth on-chain contract on a target blockchain. The contract verifies the data’s authenticity (via signatures/confidence intervals) and stores it for use.

•  Integration flow: Developers integrate Pyth in two parts — off-chain code fetches the latest price update (via API/Wormhole for cross-chain), then includes it in their on-chain transaction (e.g., a perp trade or liquidation). This bundles the update with the user’s action, paying a small fee only when needed.

•  Key advantages:

•  Cost efficiency — No constant gas burn for unused updates; fees shift to demand-driven users, enabling massive scalability and lower operational costs (up to 50% savings in high-frequency scenarios).

•  Low latency & freshness — Users get the most recent price exactly when executing (sub-second possible), ideal for high-volatility derivatives/perps trading.

•  Scalability — Supports 500+ feeds across 50+ chains without oracle-side spam/cost overload.

•  Trade-offs: Requires active management (e.g., ensuring an update is fresh in the tx), potential minor added complexity vs. always-available push feeds, but many protocols use relayers/bundlers to automate.

Bottom line: Pyth’s pull model flips the economics — making high-frequency, accurate financial data affordable and performant for DeFi (especially perps/lending on fast chains like Solana/EVM), while push models prioritize constant readiness at higher baseline cost. It’s a core reason Pyth dominates in derivatives transaction volume. DYOR; oracle designs evolve with chain tech.#Write2Earn