Pyth Network is a first-party oracle protocol that publishes real-time price data for crypto assets, equities, forex, and commodities. It's the dominant price feed on Solana and has expanded to 40+ blockchains.
If you use any Solana DeFi protocol — Kamino, Drift, Marginfi, Solend — you're relying on Pyth prices without knowing it.
What is an Oracle?
Smart contracts can't access external data on their own. They only know what's on-chain. An oracle is a service that brings off-chain data (like asset prices) on-chain in a trustworthy way.
DeFi protocols need price oracles for:
- Lending markets: knowing when to liquidate undercollateralized positions
- Perpetuals: pricing contracts against spot
- Options: computing fair value
- Synthetic assets: pegging to real-world prices
Bad oracle data = bad liquidations, exploits, and cascading failures. Oracle design is one of the most critical security decisions in DeFi.
How Pyth Works
First-party data model: Pyth sources prices directly from publishers — institutions like Jane Street, Jump Trading, Two Sigma, Cboe, and 90+ others. Each publisher reports their own observed price and confidence interval.
Aggregation: Pyth aggregates publisher prices using a weighted median. The confidence interval represents the spread across publishers. High confidence = tight agreement. Low confidence = publishers disagree, usually during volatility.
Publishing frequency: Pyth updates prices every 400ms on Solana — faster than most blockchains can even process transactions. On EVM chains, Pyth uses a pull model where prices are published on-demand.
Pull vs Push:
- Solana: Pyth pushes prices on-chain continuously
- EVM (Ethereum, Base, Arbitrum): Pyth uses pull oracle — protocols fetch the price when they need it by including a signed price update in the transaction. This is more gas-efficient.
Pyth vs Chainlink
| Feature | Pyth | Chainlink | |---------|------|-----------| | Data source | First-party (institutions) | Node operators (aggregated) | | Update frequency | 400ms (Solana), on-demand (EVM) | Every heartbeat or deviation | | Coverage | Crypto, equities, forex, commodities | Primarily crypto + some others | | Primary chain | Solana | Ethereum | | Architecture | Publisher → aggregation → on-chain | DON (Decentralized Oracle Network) |
Chainlink uses a network of node operators who each pull price data from multiple sources. Pyth uses direct first-party data from the institutions that actually make markets in these assets.
Neither is universally better. Chainlink has deeper Ethereum integration. Pyth has lower latency and first-party data quality.
Pyth on Multiple Chains
Pyth launched on Solana but now publishes to Ethereum, Base, Arbitrum, Optimism, BNB Chain, Aptos, Sui, and many others via the Wormhole bridge. A single cross-chain price feed backed by the same publisher network.
For SovereignSwap's signal layer: price data from Pyth feeds into the AI signal computation, giving the model real-time, high-confidence price data across assets.
PYTH Token
PYTH is Pyth's governance token. Holders vote on protocol parameters: which assets to support, publisher onboarding, fee structures.
Pyth has run airdrop campaigns for DeFi users, distributing PYTH to wallets that used Pyth-integrated protocols. The token trades on major exchanges and Solana DEXes.
Oracle Risk in DeFi
Even Pyth isn't risk-free:
- Publisher failure: if a major publisher goes offline, aggregate confidence widens
- Stale prices: network congestion can delay updates, causing stale price liquidations
- Cross-chain latency: on EVM pull model, a price update included in the wrong block could be stale
Protocols mitigate this with staleness checks (reject prices older than X seconds) and confidence bounds (reject if confidence is too wide). Understanding oracle risk matters when using any DeFi protocol at size.