PythFeeds is a real-time price dashboard powered by Pyth Network oracle data. It streams live prices for 600+ assets — crypto, US equities, metals, commodities and forex — with confidence intervals and sub-second updates.

How is PythFeeds different from CoinGecko or CoinMarketCap?

PythFeeds surfaces oracle-grade data: every price comes with a confidence interval (price ± conf) and a publish timestamp for staleness detection, sourced from institutional publishers (exchanges, market makers). Aggregator sites combine tier-2 APIs; PythFeeds shows the same data used by on-chain smart contracts.

What is a Pyth confidence interval?

A Pyth confidence interval is a bound on how certain the oracle is about a price. Real price is approximately price ± conf. For lending, conservative practice uses price − conf for collateral and price + conf for debt, to stay safe against price uncertainty.

Is PythFeeds free to use?

Yes. PythFeeds is free. Price data is sourced from the public Pyth Network Hermes API. Solana swaps execute on-chain via Jupiter; PythFeeds never custodies user funds.

What blockchains does PythFeeds swap on?

Swaps execute on Solana via the Jupiter aggregator. Price validation uses Pyth Hermes feeds to ensure trades reference oracle-accurate rates.

How does the correlation matrix work?

PythFeeds computes Pearson or Spearman correlation on log returns (not raw prices — using returns removes trend bias). It also shows per-asset β vs BTC, annualized volatility σ, a diversification score, rolling correlation sparklines, and clustering order to reveal asset groupings.

Glossary

Plain-English definitions for the Pyth Network, oracle, DeFi, trading and portfolio terms used throughout PythFeeds. Bookmark any entry — every term has a stable URL like /glossary#confidence-interval.

Oracle

Confidence interval (conf): A bound Pyth Network attaches to every price indicating how certain the oracle is. The real price is approximately price ± conf. Lending protocols use price − conf for collateral valuation and price + conf for debt valuation to stay safe against uncertainty.
See also: Publish time, Pull oracle
Publish time: The UNIX timestamp marking when a Pyth price was last aggregated. PythFeeds surfaces a green/amber/red staleness pill — green under 2 seconds old, amber under 10, red beyond.
See also: Confidence interval (conf)
Pull oracle: An oracle model where prices are published off-chain and pulled on-demand by smart contracts, paying a small update fee. Pyth Network uses this model (unlike Chainlink's push model). It is more efficient and supports sub-second updates.
See also: Push oracle, Hermes API
Push oracle: An oracle that writes prices on-chain on a schedule, independent of consumer demand. Higher infrastructure cost, slower updates. Chainlink is the canonical example.
See also: Pull oracle
Hermes API: The off-chain Pyth Network HTTP and SSE service (hermes.pyth.network) that distributes signed price updates. PythFeeds streams prices directly from Hermes.
See also: Pull oracle
Feed ID: A unique 32-byte identifier for each Pyth price feed. For example, ETH/USD is 0xff61491a931112ddf1bd8147cd1b641375f79f5825126d665480874634fd0ace.

Trading

Slippage: The difference between the expected swap price and the actual fill. PythFeeds lets you set a tolerance (0.1% / 0.5% / 1% / custom) and flags quotes that deviate further than allowed from the Pyth oracle price.
Limit order: A conditional order to trade at a target price or better. PythFeeds monitors Pyth prices and fires a Jupiter swap when the condition is met.
TWAP (time-weighted average price): An average price weighted by time, used to smooth out short-term noise. Smart contracts often use TWAPs to reduce oracle-manipulation risk.

Portfolio

Sharpe ratio: (Mean return − risk-free rate) / standard deviation of returns, annualized. Higher is better risk-adjusted performance. PythFeeds computes a portfolio Sharpe once you have ≥14 days of history.
See also: Sortino ratio, Max drawdown
Sortino ratio: Like Sharpe but only penalizing downside volatility. Useful when upside spikes shouldn't count as 'risk'.
See also: Sharpe ratio
Max drawdown: The largest peak-to-trough decline in a portfolio's value. A −30% drawdown means at its worst the portfolio was 30% below a prior peak.
Diversification score: On the correlation matrix, 1 − mean(|r|) across all off-diagonal pairs. 100% means zero average correlation (perfect diversification); near 0% means everything moves together.

DeFi

Depeg: When a pegged asset (usually a stablecoin like USDC, USDT or DAI) trades at a meaningful distance from its peg. PythFeeds offers depeg alerts with a configurable threshold (e.g. 0.5%).
Jupiter: The leading Solana DEX aggregator. PythFeeds routes all swaps through Jupiter and validates the quoted output against Pyth oracle prices.

Statistics

Log returns: Returns computed as log(p_t / p_{t-1}) instead of raw percentage change. Log returns are additive over time and remove trend bias, making them the correct input for correlation and volatility analysis.
See also: Pearson correlation, Spearman correlation, Volatility (σ)
Pearson correlation: A measure of linear correlation between two series, ranging from −1 (perfect inverse) to +1 (perfect co-movement). Sensitive to outliers.
See also: Spearman correlation, Log returns
Spearman correlation: Pearson correlation applied to the ranks of the observations. Robust to outliers and captures monotonic (non-linear) relationships. Preferred when markets have fat-tail days.
See also: Pearson correlation
Beta (β): The OLS regression slope of an asset's returns on a benchmark's returns (usually BTC). β = 1.5 means the asset moves 1.5× as much as BTC. β < 0 indicates inverse movement.
Volatility (σ): The standard deviation of returns, measuring how dispersed an asset's moves are. PythFeeds annualizes hourly volatility by √(24 · 365).
See also: Log returns

Trending Now

Watchlist

Glossary

Oracle

Trading

Portfolio

DeFi

Statistics

More reading

Glossary

Oracle

Trading

Portfolio

DeFi

Statistics

More reading