Pyth Network (PYTH) Explained: Pull Oracle, Lazer, and 2026 Guide

TL;DR

• Pyth Network is the dominant pull-oracle for high-frequency DeFi. By April 2026 it covers 600+ price feeds, sourced from 90+ first-party publishers (Jane Street, Jump Trading, Wintermute, Cboe Global Markets, Virtu, Two Sigma, IMC Trading), and is consumed by smart contracts on 50+ blockchains via Wormhole-relayed VAAs.
• The PYTH token launched on November 20, 2023 in the largest oracle airdrop in crypto history; it governs Pyth Improvement Proposals (PIPs) and powers Oracle Integrity Staking (OIS).
• Pyth Lazer (February 2025) is the network's sub-millisecond latency tier — 1 ms cadence and ~15,000 Solana compute units per call — powering Hyperliquid, GMX V2, Drift, Aevo and other perp DEXes.
• Pyth Express Relay brings MEV-aware execution to integrating protocols; Pyth Entropy is a commit-and-reveal verifiable randomness service that competes with Chainlink VRF.

Table of contents

• What is Pyth Network?
• Origins: Jump Trading, Solana 2021, Pyth Data Association
• The first-party publisher model
• Push vs pull: Pyth's architectural bet
• How a Pyth price update flows on-chain
• Pyth Lazer: sub-millisecond latency
• Pyth Express Relay: MEV-aware execution
• Pyth Entropy: verifiable randomness
• Pyth Data Marketplace and beyond price
• The PYTH token, airdrop and tokenomics
• Oracle Integrity Staking (OIS) and PIP governance
• Adoption: Hyperliquid, GMX, Drift, Aevo and Solana DeFi
• How Pyth compares to Chainlink, RedStone, Switchboard, Stork, eOracle
• Use cases
• Risks and criticism
• How to use Pyth as a developer
• How to become a Pyth publisher
• Research and reports
• FAQ
• Glossary
• Related reading
• Sources and further reading

What is Pyth Network?

Pyth Network is a decentralized oracle network that streams real-time financial market data — cryptocurrencies, US and international equities, ETFs, FX pairs, commodities, rates, and increasingly real-world index data — from the trading firms and exchanges that produce the data, to smart contracts on dozens of blockchains. By April 2026 Pyth Network supports more than 600 price feeds drawn from more than 90 first-party publishers and is consumed by smart contracts across more than 50 blockchains including Solana, Ethereum, Arbitrum, Base, Optimism, BNB Chain, Aptos, Sui, Polygon, Avalanche, Hyperliquid, Berachain, Sei, Sonic, Mantle, Linea, Mode, Blast, Scroll and many more.

Pyth's core architectural choice is the pull oracle: rather than pushing a new on-chain price on a deviation or heartbeat trigger (like Chainlink Data Feeds), Pyth keeps a continuously-updated, off-chain, signed price stream and lets each consuming smart contract pull the latest signed price into its own transaction. This inverts the cost model: there is no on-chain write cost when no one is trading, but each trade pays a small fee to attach the latest price update — a pattern that maps perfectly onto high-frequency perpetuals trading and is responsible for Pyth's dominance of the on-chain perps stack.

The other distinguishing feature is the first-party publisher model: the firms that produce the price data — market makers like Jane Street, Jump Trading, Wintermute, Virtu, Two Sigma, IMC Trading, Hudson River Trading, Susquehanna and exchanges like Cboe Global Markets, Binance, OKX, Bybit, Kraken, Coinbase Institutional — publish their internal prices directly, signed with their corporate keys. Chainlink, by contrast, is a third-party model: independent node operator companies fetch data from public APIs and aggregate it. The first-party model is intended to be cleaner (no API-scraping latency, no upstream API outages), more accountable (a publisher signs with a known key and stakes reputation), and more aligned with the institutional data sources that already produce the prices.

Origins: Jump Trading, Solana 2021, Pyth Data Association

Pyth was conceived inside Jump Trading's crypto arm in 2020-2021 to solve a specific problem: Jump and other quantitative trading firms had no clean way to feed institutional-quality, low-latency prices into Solana DeFi (where Jump was a leading market maker on Serum and Mango Markets). The first prototype launched on Solana mainnet in mid-2021 with a handful of feeds and a small set of Jump-affiliated publishers; the official Pyth Network mainnet went live in August 2021 with crypto, FX, equities and commodities feeds.

Stewardship was transitioned over 2022-2023 to the Pyth Data Association, a Swiss not-for-profit verein (Switzerland's traditional industry-association legal form) that holds the Pyth trademark and the governance keys, and to Douro Labs — the for-profit core contributor company that builds the protocol. Douro was co-founded in 2023 by Mike Cahill (CEO and ex-Jump Crypto Head of Capital Markets), Jayant Krishnamurthy and Ciaran Cronin, all long-time Pyth contributors. The PYTH token launched in November 2023, completing the transition from Jump-incubated project to open, DAO-governed protocol.

The first-party publisher model

The Pyth publisher list reads like a who's-who of global market-making and electronic trading. By April 2026 the network counts 90+ approved publishers including:

• Market makers: Jane Street, Jump Trading, Wintermute, Virtu Financial, Two Sigma Securities, IMC Trading, Hudson River Trading, Susquehanna International Group, GTS, Optiver, Flow Traders, Flowdesk, Auros, Galaxy Digital Trading, Cumberland (DRW), Amber Group, GSR.
• Exchanges and brokerages: Cboe Global Markets, Binance, OKX, Bybit, Kraken, KuCoin, Bitstamp, Coinbase Institutional, BitMEX, Crypto.com, Bitfinex, Bitget, Gate.io, Kraken, MEXC.
• Crypto-native firms: 0x, Alameda (legacy), Genesis (legacy), B2C2, GSR, Kairon Labs, Skynet Trading, FalconX, LMAX, FTX (legacy).

Each publisher runs signing infrastructure that pushes its internal best-bid/best-ask/last-trade prices to the Pythnet aggregation chain at 400 millisecond cadence. The aggregator (Pythnet, a Solana fork operated by the Pyth Data Association) computes a robust median across publishers, weighs each publisher's confidence interval, and emits a single canonical signed price plus a confidence interval (the "Pyth confidence band") for downstream consumption. The empirical paper How DeFi Protocols Choose Oracle Providers (December 2025) finds that Pyth's confidence-band design materially reduces oracle-induced bad-debt during volatile periods compared to a naive median.

Push vs pull: Pyth's architectural bet

The push-versus-pull distinction is the most important architectural choice in oracle design and the cleanest way to understand where Pyth fits. A push oracle continuously writes new prices on-chain whenever a deviation or heartbeat triggers, regardless of whether any contract actually consumes the price in that block. Aave's GHO redemption code reads ETH/USD whether or not it's been called recently; the on-chain price is "always there." This is great for slow-moving lending markets but expensive when many feeds are pushed across many chains.

A pull oracle keeps the price stream off-chain on a high-throughput aggregation layer and asks the consumer to attach the latest signed price as part of their transaction. The on-chain footprint is exactly proportional to usage: a quiet day on a perp DEX writes nothing; a busy day writes a price update with every trade. The latency is also better, because the off-chain layer can update at sub-second cadence — far faster than any push model can economically sustain on Ethereum mainnet.

The trade-offs are real. A pull oracle requires the consumer or its frontend to fetch the latest signed price (Pyth provides the Hermes REST/WebSocket service for this) and submit it. A pull oracle is also less convenient for "passive" use cases like a stablecoin whose collateral price needs to be readable at any time without a transaction — Pyth handles this via its "scheduled feeds" hybrid product, which periodically posts to a chain at a configurable interval (Pyth's answer to push when push is the right pattern).

The recent paper SoK: Blockchain Oracles Between Theory and Practice systematizes the design space and singles out the push/pull split as the central organising axis, with Chainlink and Pyth as canonical references for each.

How a Pyth price update flows on-chain

The end-to-end Pyth flow has six steps:

1. Publisher signing: 90+ publishers compute their internal price for, say, BTC/USD every ~400 ms, sign the update with their corporate Ed25519 key, and submit it to Pythnet (the Pyth aggregation Solana fork).
2. Aggregation on Pythnet: Pythnet runs the Pyth aggregator program that computes a robust median, a confidence interval, and a canonical signed price.
3. Wormhole attestation: every Pythnet block, the aggregated prices are attested by the Wormhole guardian network as a VAA (Wormhole Verifiable Action Approval) — the same VAA primitive Wormhole uses for cross-chain bridges.
4. Hermes distribution: Pyth runs a public service called Hermes that exposes the latest VAAs over REST and WebSocket. Frontends, keepers, MEV searchers and integrators all subscribe to Hermes.
5. On-chain pull: when a user wants to trade on, e.g., GMX V2 on Arbitrum, the frontend fetches the latest VAA for the relevant feed from Hermes and submits it as part of the user's transaction. The Pyth on-chain contract verifies the Wormhole guardian signatures and writes the canonical price to its on-chain storage.
6. Consumer read: the GMX contract reads the just-written price via getPriceUnsafe() or getPriceNoOlderThan() and uses it for execution.

The whole flow has typical end-to-end latency of 400-800 ms — bounded by the Pythnet aggregation cadence and the destination chain's block time.

Pyth Lazer: sub-millisecond latency

Pyth Lazer, launched February 2025, is the network's institutional high-frequency tier. Lazer publishes per-publisher signed updates at 1 millisecond cadence — 400× faster than the standard 400 ms — without going through the Pythnet aggregation step. The consumer chooses (or the protocol pre-selects) one or more publishers, fetches their individual signed updates, and submits them to the on-chain Lazer verifier.

Two design choices make Lazer interesting:

• Compute cost: a Lazer call costs roughly 15,000 Solana compute units versus ~30,000 for the standard Pyth call. Across thousands of trades per second, the saving compounds into meaningful rebates.
• Latency: 1 ms publisher cadence means Lazer's latency is bounded almost entirely by network propagation and the destination chain's block time. On Solana that's effectively the ~400 ms slot time; on Hyperliquid's L1 it's the consensus latency of ~70 ms.

Lazer's launch design partners include Hyperliquid (the leading on-chain perp DEX, with cumulative volume in excess of US$2T by 2026), Drift Protocol, and Aevo. It is the answer to the criticism that public-blockchain oracles couldn't match the latency of CME or Eurex order books — Lazer narrows the gap to a single-digit milliseconds.

Pyth Express Relay: MEV-aware execution

Pyth Express Relay, launched 2024, is the protocol's MEV-recapture and best-execution venue. Conceptually it's the application-layer cousin of MEV-Boost on Ethereum.

The mechanism: a DeFi protocol that integrates Express Relay routes user transactions through a private mempool. Searchers bid (in PYTH or stablecoin) for the right to be the executor of the user's order. The winning searcher pays the bid; the protocol captures most of the bid as MEV recapture; the user gets best execution and pays a single dynamic fee that is typically lower than the implicit cost of public-mempool sandwiching.

Express Relay is integrated by lending protocols (for liquidation backruns), perp DEXes (for fill optimization), and DEX aggregators. It runs across 30+ chains and is settlement-agnostic — Pyth attaches no opinionated chain-specific assumptions, so an integration that started on Solana can re-deploy to Arbitrum or Base with minor wiring changes.

Pyth Entropy: verifiable randomness

Pyth Entropy is Pyth's on-demand verifiable randomness primitive, designed as a head-to-head competitor to Chainlink VRF. Entropy uses a commit-and-reveal mechanism rather than VRF's signed-elliptic-curve approach:

1. The consumer requests randomness, providing a user-supplied "seed".
2. A network of providers committed in advance to a hash-chain of pre-images; the Entropy contract reveals the next pre-image.
3. The verifier checks the pre-image against the committed hash and combines it with the user seed (and optionally the destination block hash) to produce final randomness.

The advantage over VRF: no on-chain elliptic-curve verification, which is expensive on EVM chains. The trade-off: the commit-and-reveal model is sensitive to provider liveness — a provider that fails to reveal can stall a request, requiring fallback to alternate providers. Entropy is live on Ethereum, all major L2s, Solana, Aptos, Sui, and 30+ others, and is widely used for NFT minting, gaming, lottery and validator selection.

Pyth Data Marketplace and beyond price

Pyth Data Marketplace (sometimes called Pyth Pro) is the network's expansion beyond simple price feeds. It includes:

• Benchmarks: historical TWAPs, VWAPs and reference rates for institutional reporting and on-chain settlement.
• Yield curves: tokenized fixed-income reference data.
• Forex fixings: exchange-grade FX fixings (mid, bid, ask) for tokenized FX products.
• Custom feeds: bespoke feeds for institutional clients covering rates, indices, commodities and tokenized RWA NAVs.

The strategic ambition is to make Pyth the on-chain version of Bloomberg or Refinitiv — the canonical source of financial reference data for tokenized capital markets — rather than only a real-time price oracle for DeFi.

The PYTH token, airdrop and tokenomics

PYTH is the governance token of the Pyth Network. It launched on November 20, 2023 in what is widely regarded as the largest oracle airdrop in crypto history — approximately 90,000 wallets received an initial 255M tokens (~2.55% of the 10B max supply), with subsequent retroactive distributions in 2024 and a publisher-incentive distribution in 2025. The Pyth Airdrop drew millions of claim transactions on Solana and crashed several portfolio aggregators on launch day.

PYTH tokenomics:

PYTH is primarily a Solana SPL token with bridged versions on Ethereum and other chains via Wormhole. By April 2026 PYTH is in the top 80 cryptocurrencies by market cap, with a fully-diluted valuation in the multi-billion-dollar range.

Oracle Integrity Staking (OIS) and PIP governance

Oracle Integrity Staking (OIS), launched 2024, is Pyth's economic-security mechanism. PYTH holders delegate stake to specific publishers; the delegated stake earns a share of network fees that publisher generates, and is slashed if that publisher's data is provably inaccurate or stale during a misbehaviour event. OIS aligns publishers, delegators and the protocol — publishers benefit from delegated stake because it boosts their rewards, delegators do due diligence on publisher quality, and the protocol gains a financial fault-tolerance layer beyond pure reputation.

Pyth Improvement Proposals (PIPs) are the on-chain governance vehicle. Holders of staked PYTH (in the OIS or governance staking pool) vote on PIPs that cover fee parameters, publisher onboarding/offboarding, treasury allocations, and protocol upgrades. PIPs are administered through a Solana-native governance contract and have averaged ~10-20 active proposals per quarter through 2025-2026.

Adoption: Hyperliquid, GMX, Drift, Aevo and Solana DeFi

Pyth's adoption pattern is the mirror image of Chainlink's. Where Chainlink dominates DeFi lending and stablecoin reserves, Pyth dominates high-frequency perpetuals, Solana DeFi, and the on-chain orderbook stack.

Hyperliquid's adoption is the most economically significant: by 2026 the venue routinely processes US$10-20B in daily perp volume and has cumulative volume well in excess of US$2T. Hyperliquid's core perp mark prices are set by its own validator-run oracle rather than by Pyth; Pyth's role on Hyperliquid is concentrated in HIP-3 builder-deployed markets.

How Pyth compares to Chainlink, RedStone, Switchboard, Stork, eOracle

The 2026 oracle landscape is well-differentiated:

Many of the high-volume protocols use two oracles — a primary Pyth + a Chainlink safety bound (e.g., GMX V2), or a primary Chainlink + a Pyth high-frequency layer. The empirical December 2025 study How DeFi Protocols Choose Oracle Providers finds dual-oracle configurations are now the modal design for venues processing more than US$100M of daily volume.

Use cases

Pyth's product-market fit is concentrated in a handful of demanding use cases:

1. High-frequency perpetuals: Hyperliquid, GMX V2, Drift, Aevo, Synthetix, Jupiter Perps. Sub-second latency and confidence-band-aware liquidation logic make Pyth the dominant choice.
2. Solana DeFi: Kamino, MarginFi, Mango, Save, Jito, Sanctum, Drift, Phoenix, OpenBook all rely on Pyth as their primary oracle. Solana's native account model integrates cleanly with Pyth's pull contract.
3. Cross-chain expansion of Solana DeFi: as Solana-native protocols expand to Ethereum and L2s (e.g., Drift's Aevo cross-listing, Jupiter expanding to EVM), Pyth follows them and provides continuity.
4. Tokenized FX and equities: tokenized stocks (Backed bShares, Dinari dShares, Ondo Global Markets) increasingly use Pyth FX and equity feeds for redemption, NAV calculation and secondary-market mid-pricing.
5. MEV recapture (Express Relay): lending protocols and DEXes capturing the MEV that previously leaked to public-mempool searchers.
6. Verifiable randomness (Entropy): NFT mints, gaming, lottery, validator selection — Entropy is materially cheaper than Chainlink VRF on EVM chains.
7. Reference data for tokenized RWAs: Pyth Benchmarks for TWAPs and yield curves used by tokenized treasury issuers and on-chain repo venues.

Risks and criticism

No oracle is risk-free. The most-cited Pyth critiques:

• Solana dependency: Pyth's aggregation chain (Pythnet) is a Solana fork. Solana network outages (most notably September 2021 and 2022) historically risked Pyth liveness. Recent upgrades (independent Pythnet validators, Wormhole VAA caching) have largely mitigated the dependency, but the architecture is more Solana-coupled than Chainlink's chain-agnostic OCR2 design.
• Wormhole counterparty risk: Pyth distribution is anchored to Wormhole VAAs. The infamous February 2022 Wormhole Bridge $325M hack was on Wormhole's bridge contract, not on its messaging core, but the brand association leads some integrators to demand additional safety bounds — typically in the form of a Chainlink secondary feed.
• Confidence-band misuse: Pyth's confidence band must be correctly handled by the consumer. Several documented bad-debt incidents on Solana lending venues in 2022-2023 traced back to consumers ignoring or mis-implementing the confidence-band check during volatile windows.
• Publisher concentration in some feeds: while Pyth has 90+ approved publishers, individual feeds may have only 8-15 active publishers. A long-tail or newly-listed feed can be vulnerable to publisher coordination, mitigated by OIS slashing.
• Token unlocks: PYTH unlocks at 6/18/30/42 month cliffs from launch (Nov 2023). Each cliff materially expands circulating supply.
• Token utility critique: PYTH's economic role is governance + staking, but the network's fee revenue is still relatively small compared with the aggregate market cap, making "value-accrual to PYTH" a narrative that depends on multi-year revenue growth from Express Relay, Lazer and Data Marketplace.

The same set of price-manipulation papers cited in the Chainlink analysis apply broadly to Pyth (AiRacleX, LLM-Powered Detection of Price Manipulation, DeFiGuard, Price manipulation schemes of new crypto-tokens in DEXes, Detecting Various DeFi Price Manipulations) and confirm that pull oracles with diverse first-party publishers and well-handled confidence bands are among the most attack-resistant designs.

How to use Pyth as a developer

Practical integration steps:

1. Pick your destination chain. Pyth supports 50+ chains including Solana, Ethereum mainnet, Arbitrum, Optimism, Base, BNB Chain, Avalanche, Polygon, Aptos, Sui, Hyperliquid L1, Sei, Sonic, Mantle, Linea, Mode, Blast, Scroll. Check docs.pyth.network for per-chain contract addresses.
2. Pick your feeds. Pyth's feed catalogue covers crypto pairs, US and international equities and ETFs, FX majors and minors, commodities (gold, silver, oil), rates (SOFR, EFFR) and benchmarks. Each feed has a unique 32-byte ID.
3. Choose Lazer or standard. Lazer for HFT-grade applications (perp DEXes, low-latency arbitrage); standard Pyth for everything else.
4. Install the SDK. @pythnetwork/pyth-evm-js or @pythnetwork/client for Solana, plus on-chain Solidity / Anchor crates.
5. Fetch from Hermes. In your frontend or keeper, call the Hermes REST or WebSocket service to retrieve the latest VAAs.
6. Submit on-chain. Call IPyth.updatePriceFeeds(updateData) with a small fee in native gas; then read the price via IPyth.getPriceUnsafe(priceId) or IPyth.getPriceNoOlderThan(priceId, age).
7. Handle the confidence band. Always verify the confidence field is below your protocol's tolerance — typically 0.5% of the price for non-stable assets — and reject the trade or fall back to a secondary oracle if it's exceeded.
8. Test on devnet/testnet. Solana devnet, Ethereum Sepolia, and most other chains have Pyth deployments. Free test feeds are exposed via Hermes.

How to become a Pyth publisher

If you operate a market-maker, exchange, or institutional trading firm and have proprietary price data, you can apply to become a Pyth publisher.

1. Apply via the Pyth Data Association. Submit institutional documentation, trading-volume references, signing-infrastructure plans, and the list of feeds you intend to publish.
2. PIP approval. Existing PYTH stakers vote on a Pyth Improvement Proposal that whitelists your publisher key and assigns initial feeds.
3. Stand up signing infrastructure. Run the Pyth publisher binary on dedicated infra co-located near your trading systems. Expected SLA: 99.9% uptime, <100ms publishing latency.
4. Earn fees + delegated stake. Approved publishers earn a share of network fees in proportion to their rewards eligibility, plus optional delegated PYTH stake from OIS — and accept slashing risk on data quality. Top publishers earn 7-figure annualized revenue from network fees alone in 2026.

Research and reports

Suggested reading on the academic and industry literature:

• SoK: Blockchain Oracles Between Theory and Practice (ePrint 2025/2106) — definitive recent oracle survey.
• How DeFi Protocols Choose Oracle Providers: Evidence on Sourcing, Dependence, and Switching Costs — empirical study including Pyth dual-oracle adoption.
• Ormer: A Manipulation-resistant and Gas-efficient Blockchain Pricing Oracle for DeFi — manipulation-resistance for pull and push designs.
• AiRacleX: Automated Detection of Price Oracle Manipulations — ML-based oracle attack detection.
• Detecting Various DeFi Price Manipulations with LLM Reasoning — companion ML defense paper.
• LLM-Powered Detection of Price Manipulation in DeFi — late 2025 survey.
• DeFiGuard: A Price Manipulation Detection Service in DeFi using Graph Neural Networks — graph-network manipulation detection.
• Price manipulation schemes of new crypto-tokens in decentralized exchanges — long-tail listing manipulation.

FAQ

What is Pyth Network and how does it work?

Pyth is a pull oracle that streams real-time financial data from 90+ first-party publishers (Jane Street, Jump, Wintermute, Cboe, Virtu, Two Sigma, IMC, etc.) to the Pythnet aggregation chain at 400 ms cadence; consuming smart contracts on 50+ chains pull the latest signed price into their transaction via Wormhole VAAs. 600+ feeds in 2026.

Who founded Pyth Network?

Incubated at Jump Trading in 2020-2021; launched on Solana in 2021; spun out under the Pyth Data Association (Swiss verein) and Douro Labs, co-founded in 2023 by Mike Cahill (CEO), Jayant Krishnamurthy and Ciaran Cronin.

What is the difference between push and pull oracles?

Push (Chainlink) writes prices on-chain on deviation/heartbeat; pull (Pyth) keeps prices off-chain and lets contracts pull the latest signed price into their transaction. Pull is faster and gas-efficient for high-frequency consumers; push is efficient for slow lending markets.

What is Pyth Lazer?

A sub-millisecond latency tier launched February 2025: 1 ms publisher cadence, ~15K Solana compute units (vs ~30K standard), powering Hyperliquid, Drift, Aevo and other perp engines.

What is the PYTH token?

Solana SPL token launched November 20, 2023 in the largest oracle airdrop in crypto history. 10B max supply. Governs PIPs and powers Oracle Integrity Staking.

Which protocols use Pyth?

Hyperliquid, GMX V2, Drift, Aevo, dYdX v4, Synthetix, Kamino, MarginFi, Mango v4, Save, Jupiter Perps and most of Solana DeFi. Many pair with Chainlink.

What is Pyth Express Relay?

MEV-aware execution venue with private-mempool searcher auctions; conceptually parallel to MEV-Boost but at the application layer and chain-agnostic.

What is Pyth Entropy?

Commit-and-reveal verifiable randomness — competitor to Chainlink VRF. Cheaper on EVM because no on-chain elliptic-curve verification.

How does Pyth compare to Chainlink?

Chainlink push dominates DeFi lending, RWA and SWIFT integration. Pyth pull dominates high-frequency perps and Solana DeFi. Many protocols use both. Chainlink Data Streams was the direct competitive response.

How do I integrate Pyth?

Pick chain + feeds + Lazer-or-standard, install SDK, fetch from Hermes, call updatePriceFeeds() with a small fee, read via getPriceUnsafe() or getPriceNoOlderThan(), always check the confidence band.

Glossary

• Pull oracle — design where prices live off-chain and consumers pull signed updates into their transaction.
• First-party publisher — a market maker or exchange that signs and publishes its own internal prices to the oracle.
• Pythnet — Pyth's Solana-fork aggregation chain.
• VAA — Wormhole Verifiable Action Approval, the cross-chain message format Pyth uses for distribution.
• Hermes — Pyth's REST/WebSocket service exposing the latest VAAs to integrators.
• Confidence band — Pyth's per-update interval that consumers must check before acting on a price.
• Lazer — Pyth's sub-millisecond per-publisher latency tier.
• OIS (Oracle Integrity Staking) — Pyth's economic-security mechanism where PYTH holders delegate stake to publishers and share rewards plus slashing risk.
• PIP (Pyth Improvement Proposal) — on-chain governance proposal voted on by staked PYTH.
• Express Relay — Pyth's MEV-aware private-mempool execution venue.
• Entropy — Pyth's verifiable randomness primitive.

Related reading

• Chainlink (LINK) Explained: Oracles, CCIP, and SmartCon 2026
• LayerZero (ZRO) Explained: Omnichain Messaging and 2026 Guide
• What is DeFi? — 2026 Guide
• What is Ethereum? — 2026 Guide
• Real-World Asset Tokenization: 2026 Guide
• Ethereum Layer 2 Networks — 2026 Guide
• Stablecoins Explained — 2026 Guide

Related comparisons

• Compare: Chainlink vs Pyth
• Compare: Wormhole vs Axelar
• Compare: Wormhole vs Hyperlane

Sources and further reading

• Pyth Network — https://pyth.network
• Pyth Whitepaper — https://pyth.network/whitepaper
• Pyth Documentation — https://docs.pyth.network
• Pyth Blog — https://pyth.network/blog
• Pyth Data Association — https://www.pythdataassociation.com
• Douro Labs — https://www.douro.com
• Wormhole — https://wormhole.com
• DeFiLlama Oracles dashboard — https://defillama.com/oracles
• Hyperliquid — https://hyperliquid.xyz
• GMX — https://gmx.io
• Drift Protocol — https://drift.trade
• dYdX — https://www.dydx.exchange
• Cboe Global Markets — https://www.cboe.com
• Virtu Financial — https://www.virtu.com
• Jane Street — https://www.janestreet.com
• Jump Crypto — https://www.jumpcrypto.com
• IACR ePrint 2025/2106 (SoK: Oracles) — https://eprint.iacr.org/2025/2106
• arXiv 2512.03088 (Oracle Sourcing) — https://arxiv.org/abs/2512.03088
• arXiv 2410.07893 (Ormer manipulation-resistance) — https://arxiv.org/abs/2410.07893
• Pyth Lazer announcement (Feb 2025) — https://pyth.network/blog/lazer

About the author

DeFi Intel Research is the institutional research arm of DeFi Intel, a digital-asset trading firm. Our analysts cover oracle networks, MEV, cross-chain interoperability and on-chain market-microstructure. We integrate Pyth pull feeds and Lazer in our high-frequency execution stack, run a Solana validator, and publish quarterly post-mortems on oracle-related bad-debt events. About the author.

Last updated: 2026-04-26