Pyth Explained for Kalshi and Polymarket Traders

Traders on prediction market platforms such as Kalshi and Polymarket have been seeing a new kind of market open up, one that is more technical than your typical prediction market. These markets ask questions like, “Will gold go up or down in the next hour?”, “Will silver close above X by the end of the day?”, “Will this ETF close higher than its open?”, and “Will this stock close above Y strike price?” All of these markets are clearly faster and more technical than typical prediction contracts such as “Will the Fed cut rates?” or “Will X win?”

Many of these markets are powered by Pyth.

To start, Pyth is not a prediction market. It is not a betting exchange. It is not a competitor. It is an oracle network that feeds real-time market prices to other networks. If you have an application and you want to quickly get accurate market prices for your assets, Pyth can do that for you. Kalshi and Polymarket use Pyth feeds to set and resolve certain contracts. You can find out more about the Pyth Pro price feed and why Polymarket uses it on the Pyth Network website.

Below is an overview of what Pyth is, how it works in the prediction market context from a trader’s perspective, why prediction markets use Pyth, how it affects the markets you can trade, and how it makes trading prediction markets feel a lot like high-frequency trading.

What Is Pyth?

Simply put, the Pyth Network is a decentralized oracle network. An oracle is a network that provides data from the “real world” to blockchain applications. The real-world data that Pyth provides is financial market prices, including prices for cryptocurrencies, foreign exchange (forex), commodities, stocks, exchange-traded funds (ETFs), and other assets. You can read the Pyth Network blog for more details on Kalshi’s choice of Pyth as a prediction-market data standard.

The Pyth Network gets prices from more than 100 first-party price providers. These first-party providers are not random price scrapers. They are actual exchanges, market makers, and other large trading firms with direct access to live market prices. Each price provider submits a price and its own confidence interval, which is an assessment of how uncertain it thinks the price is.

The Pyth Network then aggregates the data from each price provider into one aggregated price and confidence interval. It publishes this data every 400 milliseconds, approximately 2.5 times per second.

This means that when you are trading a Pyth-powered market on a platform such as Polymarket or Kalshi, you may not simply be betting on what you think will happen. You may be betting on what the Pyth feed said at the resolution time.

This is a big deal.

Why the Pyth Feed Matters to Traders

Let us look at an example of a Polymarket contract: “Gold up or down in the next hour?” To trade this contract, the platform needs a reference price for gold. This prevents arguments between traders and the platform over whether the gold price exceeded the strike at the expiry time.

Pyth provides that reference price.

Polymarket uses the Pyth Pro price feed for markets in commodities, major index ETFs, and US stock markets. Because of the nature of these markets, they rely heavily on Pyth’s ability to stream prices in real time through WebSocket. Kalshi has similarly chosen Pyth for prediction-market products with exposure to commodities.

Depending on the contract, Pyth-powered markets may be structured like this:

The contract resolves to Yes if the Pyth price for gold was greater than the strike price at expiry.
The contract resolves to No if the Pyth price for gold was below the strike price at expiry.
The contract resolves according to a separate rule if the Pyth price for gold was exactly equal to the strike price at expiry.

This means you are not trading on the general concept of gold’s price. You are trading on the Pyth price of gold at a specific time.

That is not necessarily good or bad, but it is definitely worth noting.

Why Did Kalshi and Polymarket Use Pyth?

If you are a prediction market offering contracts with exposure to the prices of underlying assets, you need three things:

A price that updates quickly.
A price that is hard to manipulate.
A clear settlement rule that can be settled automatically.

Pyth was designed to solve these problems.

Unlike traditional prediction-market products, where resolution is determined by humans, such as “Candidate X won the election,” “Team X won the match,” or “Company X announced the deal,” asset-price markets change every second. Human resolution works well for elections, sports, and news, but it is a poor fit for questions like, “Did Bitcoin cross $100,000 at 3 PM?”

Pyth provides:

A continuous stream of timestamped prices.
An aggregated price feed based on data from many providers rather than a single exchange or data source.
A confidence interval that shows how much uncertainty surrounds the price at that moment.

This redundancy helps prevent a poor quote from one exchange, an outage at one exchange, or a manipulated trade from affecting the feed.

The confidence interval measures how far the price may deviate from its reported value. The band is wider when providers do not agree on the price or when there is high volatility. It becomes narrower when providers mostly agree.

As a trader, interpret the price as the number and the confidence interval as a measure of how solid that number is at that moment.

How the Aggregation Works

You do not need to be a quantitative analyst to understand how this works, but it is useful to know.

Every publisher submits:

A price.
A confidence interval.
Sometimes, bid and ask data.

Pyth takes all these submissions and aggregates them into a single feed. It uses a median-style calculation, which prevents one bad quote from pushing the price too far in one direction. The confidence interval is based on how much the publishers’ inputs disagree.

The key idea is redundancy.

If an exchange reports an errant quote because of a glitch, a thin order book, or a short outage, Pyth does not take that quote at face value. It checks data from many publishers. The feed remains available as long as enough publishers are online, and it is designed to remain accurate as long as most active publishers are accurate.

This is crucial for resolving prediction markets. Users do not want a contract to resolve incorrectly because of one bad tick from a single provider.

Pyth Versus Chainlink and Traditional Oracles

Users of crypto apps may be familiar with Chainlink, a competing oracle network. Pyth is structured differently from Chainlink.

Traditional oracle networks often update only when the price moves past a certain limit or at fixed intervals. This is efficient, but it can create a time gap between the on-chain price and the live market price.

Pyth is designed for high-frequency updates. It sends price and confidence-interval data approximately every 400 milliseconds. This is important for markets where timing matters, such as hourly up-or-down markets, daily close markets, and short-term directional bets.

There is a tradeoff.

Slower oracles can be easier for casual users who do not want a market flipping back and forth because of a five-cent move. Faster oracles allow platforms to settle markets more precisely, but they also mean a market can flip at the last moment. Anyone who has watched a Yes position at 99 cents fall apart in the final minute knows the pain.

What This Means on Polymarket

Polymarket was originally built around event markets covering elections, geopolitics, crypto, culture, and current affairs. These markets are still around, and they remain a core part of the platform.

Polymarket has also extended into traditional asset markets using Pyth Pro. These include commodities, major index ETFs, and single-name US equities. With a continuous Pyth feed, Polymarket does not need to wait for a news event to resolve certain markets.

This opens up new market types:

Hourly up-or-down markets.
Daily close markets.
Price threshold markets.
Short-term directional bets on gold, silver, oil, indexes, and stocks.

For example, if a market asks, “Will gold close above this number?”, the resolution will be based on the contract rules and the Pyth price at the relevant time, not on what your broker’s chart shows.

This is powerful, but it changes the nature of the product.

Political markets are about research, good judgment, and patience. Gold markets are about speed, volatility, and sometimes luck. You can have the right idea and still lose money because the price briefly moved against you.

What This Means on Kalshi

Kalshi markets are contracts on a regulated exchange, so they are more formal than markets on many other platforms. The exchange has to ensure that contract terms are clear enough to avoid disputes and meet regulatory requirements.

Pyth allows Kalshi to settle markets that reference real-world prices without relying on a single exchange’s closing price. For example, Kalshi uses Pyth oracles to settle commodity prediction contracts tied to gold, crude oil, grains, and other assets.

For you as a Kalshi trader, the message is simple: read the resolution source.

If the market states that Pyth is the resolution source, then the outcome is based on Pyth. It will not be based on what you see in your brokerage app, TradingView, Coinbase, or a price posted on X. The contract rules will explain the exact Pyth feed, the resolution time, and any rounding or tie-breaking rules.

This may sound obvious, but it is a major difference between casual prediction and asset-price markets.

Why Pyth Markets Feel Harder

Pyth makes prediction markets more precise, but more precision is not always a good thing.

A real-time price feed introduces uncertainty that is always one tick away. A contract that seems safe at 97 cents can resolve against you if a 0.05% price move occurs in the last minute. That is not necessarily bad design. It is just a different design.

Event markets are about narratives:

The incumbent underperformed in early results.
The star striker is injured.
The deal will probably be announced this week.

Price markets are about mechanics:

Gold is 40 cents below the strike price.
Volatility is rising.
The feed updates every 400 milliseconds.
Market makers are hedging on another venue.

Both types of markets can be profitable. But only event markets feel like predicting the future. Price markets can feel like trying to beat a machine that always sees the market first.

The Casual Trader Problem

Many casual users come to Kalshi or Polymarket simply to express an opinion. They have a day job. They do not want to watch charts all day. They want to buy Yes on a topic they know something about, go back to work, and check the result later.

Pyth markets disrupt that casual model.

A gold market that resolves every hour is not about patience. It is about attention. A daily close market is more suitable for casual users, but volatility near the close can still be intense. An hourly gold market is basically a coin flip, with slightly better information for people watching the feed.

Some traders argue that Pyth has “ruined” prediction markets. But Pyth does not decide market outcomes. The venue does. The venue sets the contract terms and resolution logic. Pyth simply enables the venue to offer fast, price-based contracts, which can attract bots, market makers, and professional traders.

This can create an environment that feels hostile to casual bettors who are just looking for fun.

Is Pyth Bad for Prediction Markets?

Not necessarily. It depends on what you want prediction markets to be.

Pyth is great if you want a full financial marketplace with contracts on many assets and time frames. It is excellent infrastructure for automated, transparent, high-frequency settlement.

If you are primarily interested in belief markets, meaning markets based on opinions about real-world events, Pyth price markets can feel like a distraction. They bring liquidity and volume, but they also bring a different type of trader and a different type of risk.

The better way to look at it is this: Pyth is neither good nor bad. It is a tool.

A hammer can build a house or break a window. Pyth can make settlement more reliable, or it can turn a prediction market into a latency-sensitive playground for traders. How it is used is up to the platform.

How to Trade Pyth Markets Without Becoming a Quant

You do not need to understand order-book microstructure or write a bot. But it is worth changing your approach.

Read the Rules First

Before you trade, make sure you understand the resolution source and terms:

Which Pyth price feed does the market use?
What is the exact timestamp for resolution?
Does the market resolve using the price, bid, ask, or another value?
What happens when the price is equal to the strike?
Does resolution happen at a particular second, minute, or the close of the day?

Even if you think gold is overvalued, oil is set for a big rally, or you have a favorite approach to S&P 500 markets, the specifics of the contract matter most.

Prefer Longer Time Frames

If you want a more relaxed trading experience, avoid one-minute and one-hour markets. Daily, weekly, or event-based markets give you more room to correct early mistakes.

A daily-close gold market is still risky, but it is nowhere near as stressful as an hourly market, where you can lose over one bad minute.

Size Down

Price markets can see sharp reversals. Treat them as short-term speculation rather than long-term conviction trades. If you would not be happy losing the entire position, it is too big.

Avoid the Final Minutes

Holding a position close to the strike price during the final few minutes can be brutal. Volatility tends to increase, bid-ask spreads often widen, and a tiny price movement can flip the market.

You do not need to watch a market second by second, but in short-duration markets, most of the action happens at the finish.

Separate Fun Money From Serious Capital

If you are in it for fun, set an entertainment budget. Pyth markets can be enjoyable at small stakes. Stress sets in when every position feels important enough to monitor constantly.

The Broader Perspective

From a broader perspective, Pyth is part of a bigger change in prediction markets. Markets are moving from predicting whether an event will happen to predicting what a price will be at a particular time.

The benefits include:

Automated resolution.
Clear timestamps.
Continuous markets.
More asset classes.
Less reliance on manual reporting.

The downsides include:

More bot competition.
More short-term noise.
More emphasis on speed.
Less room for narrative-based judgment.
More stress for casual users.

Pyth’s technology is impressive. It aggregates data from a large number of institutional sources to deliver high-frequency price quotes and confidence intervals. For marketplaces, it provides a reliable way to resolve contracts based on asset prices.

Yet it does not necessarily make life easier if you are just trying to relax after work and place a small bet.

In Summary

If you are a Kalshi or Polymarket user, think of Pyth as the behind-the-scenes scoreboard for an increasing number of markets.

For event markets, that scoreboard is an election, a sports game, or an official announcement. For Pyth markets, it is a live price feed that updates multiple times per second.

This is fair because it gives traders a single source of truth. But it is also unforgiving. We are no longer just predicting events. We are competing against every other person or algorithm watching the same live feed, many of which are built to react faster than we can.

For fun, choose markets where your knowledge, research, and opinions actually matter. For Pyth price markets, trade small and treat them as short-term speculation with precise terms. You do not need to become a quantitative analyst.

Prediction markets should not require a Bloomberg terminal and a caffeine addiction. Sometimes, the best trade is choosing not to play the game.

Pyth Explained for Kalshi and Polymarket Traders: How It Works