Pyth Network is an oracle: a service that takes real-world prices, stocks, crypto, commodities, currency exchange rates, and delivers them onto blockchains so that smart contracts, programs that run automatically once their conditions are met, can use them. A lending app that needs to know the current price of ETH to decide whether to liquidate a loan is a typical customer.

Pyth's approach is to go directly to the source. Rather than scraping prices from public exchanges, it gets data submitted directly by more than 138 trading firms and exchanges, the institutions actually making the trades, then combines their submissions into a single, verified price. That price feed is currently used by more than 711 applications across 114 different blockchains, publishing over 3,059 individual price feeds.

Pyth is not itself a blockchain. It runs its own specialized network, called Pythnet, purely to combine and publish these price feeds, and it delivers the results everywhere else. That distinction matters for reading this assessment: several of the framework's tests are built for general-purpose blockchains and simply don't apply to a specialized data service like this one.

PYTH is a governance token. Holding it, and locking it up through staking, gives someone a vote in the Pyth DAO, the organization that controls decisions like which data products exist, what they cost, and how the network's revenue gets used.

Some of Pyth's products charge a fee for their data, paid in ways that ultimately route back to a shared treasury controlled by that DAO. That treasury uses part of what it collects to buy PYTH tokens on the open market, a mechanism called a buyback. Burning tokens (permanently destroying them) and paying dividends are both explicitly switched off; buybacks are the one channel through which Pyth's business activity is designed to affect the value of every PYTH token in existence.

Over the trailing year, Pyth's products together brought in $3,010,824.68 in fees. That is measured cleanly: complete daily data for the entire year, with no gaps.

The shape of that money is unusual. Rather than a steady trickle from many small transactions, the data shows large single-day spikes landing near the start of nearly every month since December 2025, several exceeding $400,000 in a single day. That pattern is the signature of subscription billing: a smaller number of paying customers being invoiced on a recurring monthly cycle, rather than thousands of tiny per-use fees adding up gradually.

Almost the entire history of this money is recent. All-time fees across Pyth's full existence come to $3,106,113.20, barely more than what arrived in just the last twelve months. This is not a business with a long revenue history that recently slowed down; it is a business whose revenue only meaningfully began within roughly the past year.

Pyth offers several distinct products under one roof. The original, Pyth Core, publishes price feeds for free reading by any smart contract. A newer product, Pyth Pro, sells premium data access to paying institutional customers on a subscription basis.

Pyth Pro alone accounts for roughly 80% of every dollar Pyth has ever earned, and Pyth Pro is less than a year old. That means the healthy-looking fee numbers above describe a business that, in its current form, did not really exist a year ago. It is a genuinely new and so far unproven revenue stream, not a mature one.

This matters for how much weight to put on Pyth's economics. A young, concentrated revenue source can grow quickly. It can also shrink quickly, or depend heavily on a small number of customers who could leave. Both are live possibilities that a single year of data cannot rule out either way.

Pyth's market value, the price of every PYTH in existence multiplied together, is $333,426,683. Divide that by the trailing year's fees and Pyth trades at roughly 111 times its yearly fee income; using only the most recent three months, annualised, the multiple is a lower 59 times.

Both numbers are far from the rubric's worked example of an extreme, speculative valuation, a wireless network project that traded at roughly 87,000 times its own fees; Pyth's multiple is three orders of magnitude smaller. On the specific question of whether Pyth's price is wildly disconnected from what it currently earns, the answer is no.

That comparison should be read alongside the previous section. Because the fee base itself is new and concentrated in one product, this valuation multiple describes a young business, not an established one with a long track record behind its price.

Pyth's maximum supply is fixed at 10,000,000,000 PYTH. Of that, 72.9% is currently noncirculating, meaning it exists on paper but has not yet been released to anyone. The 27.1% that has been released splits between 10% held by insiders, 9.5% sold in a private sale before PYTH was publicly tradable, and 7.6% distributed as a public airdrop.

A single large release is scheduled for 19 May 2027: 2,125,000,000 tokens, more than a fifth of the entire maximum supply, becoming available all at once. Of that release, 212,500,000 tokens are earmarked for the team behind the project and 250,000,000 for private-sale investors; the rest is allocated to purposes not yet individually named. When this happens, the noncirculating share will fall from 72.9% to roughly 51.7% in a single event.

This is a real and disclosed feature of how PYTH works, not a hidden one. It does mean that someone holding PYTH today owns a token whose available supply is set to expand substantially, and abruptly, less than a year from now.

Pyth is governed by its DAO, and in principle any PYTH holder can participate. In practice, the computers that actually confirm and publish Pyth's price feeds are run under a system called proof-of-authority, where only a fixed, permissioned set of participants is allowed to validate at all. That permissioned set is the same group of roughly 138 trading firms who submit the price data in the first place. The people providing the raw information and the people confirming it are, structurally, the same people, with no independent layer checking their work.

Pyth's own stated identity is to "democratize this data among multiple actively contributing market participants." That is true of who contributes data. It sits less comfortably against a validator set that is closed to everyone else, and against 72.9% of the token supply not yet in public hands.

There has also been real, documented change to Pyth's original free tier. In March 2026, Pyth's DAO voted, through a proposal called OP-PIP-100, to migrate the original free product, Pyth Core, onto the same technical infrastructure as Pyth Pro, the paid product. Existing applications reading Pyth Core's price feeds were not cut off; they keep reading the same data through the same contract addresses. What did change is that free, unauthenticated access to Pyth's data feed through its public interface began requiring an API key as of 26 August 2026, and the on-chain fees Pyth Core used to charge were reduced to zero as part of the same migration. Reports describing this simply as Pyth "discontinuing" its free tier overstate what actually happened; it is more accurately described as a consolidation of two products onto one piece of infrastructure, with a new access requirement attached.