# Pyth Pro MCP Server

> Hosted MCP server that gives AI agents direct access to Pyth Pro market data, including current and historical prices, candlesticks, and funding rates.

Pyth Pro MCP Server is a Model Context Protocol server from Pyth Data Association that lets AI assistants discover market data feeds and retrieve structured prices inside their research workflows. Version 1.1 was announced in a blog post dated Oct 15, 2026. It adds one-time key configuration, a tick-level price range tool, and broader market data coverage.

## What It Is

Pyth Pro MCP exposes Pyth Pro market data as tools that MCP-compatible AI clients can call. Agents can look up feeds, retrieve current prices, historical observations, candlestick data, funding rates, market sessions, and high-frequency update channels across asset classes. The hosted server runs at an MCP endpoint, and the source code can also be run locally.

## What's New in 1.1

- **One-time key setup:** Users add their own Pyth Pro access token to their MCP client configuration instead of passing it with every request. The hosted server does not use a shared key.
- **get_price_range:** Returns every available price update in a window of up to 60 seconds, so an agent can inspect the sequence of updates around an event such as an earnings announcement.
- **Entitlement-aware get_symbols:** With a Pro key, it shows which feeds exist and flags which ones the user's plan covers or that are not live.
- **Familiar pair lookup:** Users can ask for pairs such as BTC/USD across crypto, FX, equities, and metals. The server reports when more than one match remains.
- **Broader data:** Funding rates, EMA prices, market-session information, update channels down to 50 milliseconds, interest-rate feeds, and NAV feeds.
- **Clearer errors:** Responses explain whether a feed is unavailable because of an invalid token, entitlement gap, inactive feed, missing price, or unsupported channel.

## Where It Fits in Agent Workflows

Pyth positions the server as the structured market-state source alongside web research tools. The announcement describes early exploration with Exa, where Exa handles search and research and Pyth supplies prices and candlestick data to analyze how a market responded to an event.

## Features
- Hosted MCP server for Pyth Pro market data
- One-time access token setup in MCP client configuration
- get_price_range tool returning all price updates in up to 60-second windows
- Entitlement-aware feed discovery with get_symbols
- Familiar pair lookup such as BTC/USD across crypto, FX, equities, and metals
- Funding rates, EMA prices, and market-session information
- Update channels down to 50 milliseconds
- Interest-rate and NAV feeds
- Candlestick and historical price data
- Explanatory error responses for unavailable feeds

## Integrations
Claude Code, Exa

## Platforms
API

## Pricing
Freemium — Free tier available with paid upgrades

## Version
1.1

## Links
- Website: https://www.pyth.network/blog/pyth-pro-mcp-server-1.1
- Documentation: https://docs.pyth.network/price-feeds/pro/mcp
- Repository: https://github.com/pyth-network/pyth-crosschain/tree/main/apps/mcp
- EveryDev.ai: https://www.everydev.ai/tools/pyth-pro-mcp-server