Pyth Network: The evolution from DeFi oracle to on-chain financial data infrastructure

In the traditional financial world, market data is a business with annual revenues exceeding $50 billion. Over the past 44 years, Bloomberg Terminal has locked global financial institutions into expensive subscription contracts—starting at about $27,000 per year, with a minimum two-year commitment, and requiring proprietary hardware devices. The moat of this business is not technology, but channels.

On April 9, 2026, this pattern was torn open. The decentralized oracle network Pyth Network officially launched the Pyth Data Marketplace, with the initial list of data publishers including Fidelity Investments, Euronext FX, Tradeweb, OTC Markets Group, SGX FX, and Exchange Data International—six influential institutions in traditional finance. For the first time, these institutions bypassed traditional data aggregators, directly publishing and monetizing proprietary market data on-chain.

This is not just a product launch. If viewed within the macro narrative of crypto oracles transforming from "DeFi auxiliary tools" to "financial data infrastructure," this move in April 2026 may mark the true starting point of institutional data going on-chain.



## Why Six Institutions Chose Pyth

On April 9, 2026, Pyth Network announced the official launch of the Pyth Data Marketplace through its official channels. Unlike previous oracle projects that merely provided price feeds, the core innovation of Data Marketplace is a complete "institutional data monetization framework": data publishers retain full ownership, pricing rights, and attribution rights over their data, which is directly delivered on-chain via Pyth’s cross-chain distribution network.

The first batch of data categories includes spot forex benchmarks, precious metal prices, crude oil swaps, OTC trading prices, fixed income data, and reference data sets. Almost all of this data was previously confined within closed traditional terminal systems and had never been openly circulated on the blockchain in a programmable form.

A noteworthy detail is the commercial progress of Pyth Pro. This subscription-based data product for institutions achieved over $1 million in annual recurring revenue in its first month, attracting more than 80 institutional subscribers, with about 10 new inquiries each week. While the absolute numbers are modest, as a B2B product targeting traditional financial markets, this growth signals a clear trend: institutions’ demand for on-chain data services is moving from proof of concept to actual procurement.

Meanwhile, Pyth’s penetration into prediction markets is accelerating. On April 22, 2026, Kalshi, a CFTC-regulated prediction platform, integrated Pyth data for its new commodities center, covering settlement of event contracts for eight commodities including gold, silver, and Brent crude oil. Previously, another mainstream prediction market, Polymarket, also integrated Pyth. The 24/7 trading characteristic of prediction markets, which cannot provide settlement prices after traditional exchange hours, highlights the unique value of Pyth’s pull-based model.

Connecting these events reveals a clear narrative thread: Pyth is evolving from a DeFi oracle into a comprehensive institutional-grade data distribution infrastructure.

## Technology and Model: The Logic Rebuilding of Pull Oracles

To understand why Pyth Data Marketplace can attract institutions like Fidelity, it’s necessary to revisit the technical divergence in oracle architectures.

Currently, there are two core architectures in decentralized oracles: push models and pull models. Chainlink represents the former—decentralized node networks continuously publish data updates on-chain, regardless of whether applications are actively using this data. This "broadcast" architecture’s advantage is that data is always available, suitable for scenarios like lending protocol liquidations that require immediate action. Its cost is the ongoing on-chain transaction fees, and data update frequency is limited by block times.

Pyth employs a pull model, which is logically opposite: price data is updated on the off-chain system at millisecond-level frequencies, but only when a smart contract actively requests it does the latest price get packaged on-chain. This transforms the oracle from an "all-day broadcast radio" into an "on-demand podcast"—applications only pay for the data they actually use, rather than passively bearing the cost of pushing all data to the network.

This architectural difference results in a significant cost gap. Traditional push oracles incur a Gas fee each time they push a price update; when covering hundreds of assets and requiring high-frequency updates, costs grow exponentially. Pyth’s pull model decouples price updates from on-chain writes: high-frequency updates happen off-chain, with on-chain costs incurred only when data is accessed.

This cost structure difference is decisive for institutions like Fidelity when evaluating on-chain data solutions. Institutional data—especially low-frequency but high-value categories like OTC derivatives pricing and forex swaps—would be economically unfeasible to continuously push on-chain under a push model. The pull model allows data publishers to "list" data off-chain in an aggregation layer, with consumers calling on-demand and paying per use, aligning with institutional monetization logic.

As of May 2026, Pyth has delivered over 500 low-latency price feeds to more than 50 blockchain ecosystems, with data providers including top trading firms like Jump Trading and Jane Street, as well as traditional exchanges like Cboe. Standard update latency is below 1 second, and with the new infrastructure Lazer, update frequency can be further improved to as fast as 1 millisecond.

An important background note: Pyth’s history is not heavily burdened— it did not directly compete with Chainlink at the "decentralized verification layer," but chose a differentiated path: making data quality and transmission efficiency core barriers. This approach shows clear advantages in high-frequency derivatives trading and other latency-sensitive scenarios, but in conservative financial contexts requiring multi-source cross-verification, single-source data structures still need to pass stricter audits.

## Unlock Window: Analyzing the Logic of Short-term Supply Shock

As of May 19, 2026, Gate.io market data shows PYTH token at $0.04441, up 1.79% in 24 hours, with a market cap of about $255 million, and a total supply of 10 billion tokens. Over the past year, PYTH’s price declined from around $0.138, a drop of approximately 67.86%, influenced by industry cycles and multiple large unlock events.

On May 19, the Pyth Network unlocked 2.13 billion PYTH tokens as scheduled. At the previous price, this unlock’s nominal value was about $92.46 million, representing 36.96% of the circulating supply at that time. This was one of the largest cliff unlock events in crypto in 2026.

A cliff unlock means all tokens are released at once, rather than gradually over time. There is no window for market digestion, and the influx of new circulating supply is immediate.

However, equating nominal unlock size with selling pressure is a misconception. The 2.13 billion tokens do not all flow into the secondary market. According to public allocation structures, about 1.13 billion are allocated to the ecosystem treasury; approximately 537 million are awarded to first-party institutions providing data; the rest are for protocol development and other uses.

Crucially, the treasury tokens will not immediately enter the secondary market post-unlock; their release depends on the project’s ecosystem development needs. The recipients of the data publisher rewards are institutional data providers, whose monetization behaviors are driven by their own capital management strategies, and not all will sell during the unlock window.

From a supply-demand perspective, PYTH also has an internal hedging mechanism: the PYTH Reserve auto-buyback plan. As announced in December 2025, the protocol will allocate about 33% of monthly revenue to open market buybacks of PYTH tokens. Revenue sources include Pyth Pro subscriptions, core oracle services, and Data Marketplace usage fees. The bought-back tokens are held in the PYTH Reserve and are not circulated on the secondary market. While the unlock increases circulating supply, the buyback plan creates ongoing demand, and the net effect between the two determines the actual price impact.

## Industry Competition Landscape: Not Substitution, But Diversification

In discussions about crypto oracles, whether "Pyth can replace Chainlink" is a perennial topic. But from a technical architecture and business positioning perspective, this binary framing may be insufficient.

Chainlink’s dominant position in decentralized oracles remains strong. By the end of 2025, Chainlink secured over $100 billion in total value secured. Its multi-node decentralized verification model offers unmatched security in high-stakes scenarios.

Pyth’s advantage lies in a different dimension. Its first-party data source model—data directly from exchanges and market makers, without third-party aggregation nodes—is naturally suited for high-frequency trading, derivatives pricing, and prediction markets where latency is critical. In practice, the integration of prediction markets like Kalshi and Polymarket with Pyth data for commodity contract settlement demonstrates this path’s feasibility.

The core differences can be summarized as follows:

| Comparison Dimension | Pyth Network | Chainlink |
| --- | --- | --- |
| Data Source Mode | First-party institutions directly provide | Third-party nodes aggregate multi-source data |
| Data Update Mechanism | Pull (on-demand retrieval) | Push (continuous push) |
| Core Advantage | Sub-second low latency, high frequency | Decentralized verification, security |
| Cost Structure | Pay-as-you-go, low marginal cost | Continuous update costs |
| Supported Blockchains | 50+ | ~27 |
| Price Sources | 500+ | 2,000+ |
| Main Use Cases | Derivatives, high-frequency trading, prediction markets | DeFi lending, RWA, cross-chain communication |

Source: Public project documentation and industry research reports

From the market landscape, the competition in blockchain oracles is shifting from zero-sum to layered coexistence. Chainlink secures the "consensus layer" with high security, while Pyth targets the "distribution layer" with high performance. Each builds its moat in its respective domain—this pattern is more aligned with reality than a simple replacement narrative.

## Trend Projection: From "On-Chain Data" to "Data On-Chain"

If the Pyth Data Marketplace is merely viewed as a new product launch, its significance may be underestimated. A deeper structural change is that traditional financial institutions are transitioning from "using on-chain data" to "putting their own data on-chain"—a fundamentally different participation logic.

In recent years, the intersection of institutions and crypto has mainly involved investment (buying crypto assets or investing in blockchain companies) and usage (using on-chain data as alternative data sources). The emergence of Pyth Data Marketplace signals that institutions are beginning to deploy their core data assets onto blockchain infrastructure to generate direct revenue. Multiple factors drive this shift:

First, the structural gap in data distribution channels. The global financial market data industry exceeds $50 billion annually, with highly concentrated intermediary layers. In the traditional model, trading institutions submit data to exchanges, which then resell it via data vendors to buy-side firms—an elongated, fragmented chain. On-chain direct distribution offers the possibility to compress intermediaries.

Second, the demand for tokenized assets for real-time pricing. In 2026, tokenized assets are expected to expand rapidly, with firms like BlackRock and JPMorgan moving from pilots to full deployment. On-chain trading, collateralization, and settlement of tokenized assets require real-time pricing from native data sources, which traditional data distribution pipelines struggle to provide directly to smart contracts.

Third, Pyth’s strategic transformation. In April 2026, Pyth DAO approved OP-PIP-100, establishing the phased decommissioning of the original infrastructure Pythnet within 2026, shifting focus to the new infrastructure Lazer, with Pyth Pro and Data Marketplace becoming core ecosystem products. Simultaneously, the Oracle Integrity Staking reward mechanism, as per OP-PIP-103, is gradually phased out, moving from token incentives to protocol revenue-driven economics. This transition signifies Pyth’s evolution from a token-subsidized "crypto-native" project to a "financial infrastructure" relying on real commercial income.

These three points form the underlying driver of Pyth’s current narrative. But trend projection also requires distinguishing between imagination and reality.

In an optimistic scenario, the number of institutional data publishers on Pyth Data Marketplace could grow from six to dozens, expanding data categories from forex and commodities to fixed income, credit derivatives, and macro indicators. Pyth Pro’s ARR could grow from hundreds of thousands to tens of millions of dollars. Traditional financial institutions’ acceptance of on-chain data distribution could create a positive feedback loop.

In a cautious scenario, actual on-chain data consumption by institutions remains slow, with incremental growth mainly driven by crypto-native protocols’ demand for traditional financial data, rather than proactive procurement by traditional institutions. The influx of unlocked tokens could cause temporary misalignments between market cap and protocol fundamentals, stressing the robustness of the token economic model.

In a stress scenario, large-scale unlocks combined with overall market risk aversion could lead to short-term supply-demand imbalances. Institutional data publishers’ enthusiasm may wane due to market volatility, turning Data Marketplace from a "strategic shift" into a "concept validation."

All these scenarios hinge on the premise that the market recognizes the long-term trend of on-chain restructuring of data distribution, with differences only in paths and timing.

## Conclusion

The launch of Pyth Data Marketplace is less a product iteration and more a signal of the crypto industry’s deepening move into financial data infrastructure. The participation of six institutions including Fidelity and Euronext FX provides the first verifiable anchor for the narrative of "institutional data on-chain." But turning this narrative into fundamentals requires passing key milestones: digestion of unlocked circulating supply, sustained growth in Pyth Pro revenue, and substantive increase in data consumption on Data Marketplace.

Between the long-term trend of financial data industry restructuring and the short-term game of token economics, Pyth is undergoing an evolution from a technical protocol to a business entity—and the ultimate success will depend on whether data consumption, revenue, and token value can truly create a flywheel effect.

As of May 19, 2026, Gate.io market data shows PYTH at $0.04441, up 1.79% in 24 hours, with a market cap of approximately $255 million. Market sentiment remains neutral.