Why We Built the Agent API

The crypto market generates an extraordinary volume of data every second — KOL tweets, exchange news, trade fills across thousands of wallets. For human traders, HyperX organizes this into dashboards. But a growing segment of the market does not operate through dashboards at all.

AI agents, quant systems, and automated pipelines need raw, structured, real-time data delivered programmatically — WebSocket streams to subscribe to, REST endpoints to poll, and consistent response formats to parse. Until now, getting this for the Hyperliquid ecosystem meant cobbling together unofficial sources with inconsistent formats.

Today we are launching the HyperX Agent API — a dedicated API surface for programmatic consumers. Whether you are building an AI trading agent, a sentiment analysis pipeline, or a custom alerting system, the Agent API gives you direct access to the same data that powers HyperX through clean REST endpoints and persistent WebSocket streams.

What Data Is Available

The Agent API is organized into two major categories: Data API for read-only market intelligence, and Trading API for programmatic trade execution.

Data API: Real-Time Market Intelligence

The Data API provides three core real-time streams, each available through both REST and WebSocket:

Twitter / KOL Feed — A curated, real-time stream of crypto Twitter from traders, KOLs, project accounts, and market commentators. Each tweet includes follower count, user tags, interaction type (tweet, reply, quote, repost), and media attachments. Filter by minimum followers, user tags, or screen names. This feed is entirely free — no API key required, no rate limits.

Trading News Feed — Aggregated trading news from exchanges, KOL channels, and macro sources. Each item includes the source channel, markdown text, associated trading symbols, and a link to the original. Also free with no authentication required.

Hyperliquid Fills Stream — Real-time trade fill data for any wallet address on Hyperliquid. Subscribe to multiple addresses and receive instant notifications when trades execute — coin, price, size, side, timestamp, closed PnL, fee, and taker/maker status. The foundation for wallet-tracking bots, copy trading systems, and alert pipelines.

Beyond these streams, the Data API includes a comprehensive endpoint catalog covering wallet analytics, trader metrics, market statistics, and historical data.

Trading API: Programmatic Execution

The Trading API provides full programmatic access to HyperX's copy trading and order execution infrastructure on Hyperliquid. This includes endpoints for managing copy trading relationships, executing orders, querying positions, and configuring trading parameters. Authentication uses a dedicated trading token passed via the x-hyperx-trading-token header.

The Trading API endpoint catalog is organized into logical categories — copy trading management, order execution, position queries, and account configuration — each with clearly documented request and response formats.

REST and WebSocket: Choose Your Pattern

Every data stream in the Agent API supports both REST and WebSocket access, letting you choose the integration pattern that fits your architecture.

REST endpoints are ideal for periodic polling, backfill operations, and batch processing. Query the last 24 hours of KOL tweets, fetch recent news, or pull historical fills. Responses are paginated with filtering support.

WebSocket streams are designed for real-time consumption. Connect once and data flows as it happens — tweets within seconds of posting, fills the moment they execute on-chain. Connections use a simple ping/pong keepalive mechanism.

For most AI agent use cases, WebSocket is the right choice. A sentiment agent reacting to KOL tweets cannot afford polling latency. A copy trading bot mirroring whales needs fills instantly. WebSocket streams deliver sub-second latency from event to delivery.

Who Is This For

AI Agent Developers — Feed the Twitter and news streams into your agent's context window for real-time market sentiment. The Agent API page includes five dedicated AI Skill cards with copy-paste-ready integration specifications: connection details, authentication requirements, response field descriptions, and working code examples. Copy a skill card directly into your agent's system prompt, and it can immediately start consuming HyperX data.

Quantitative Developers — Build systematic strategies with direct access to real-time order flow via the fills stream, wallet analytics endpoints, and social signal correlation. The weight-based rate limit system scales from a free tier for development to higher tiers for production workloads.

Data Analysts and Researchers — Pull structured historical data through REST endpoints for studying market microstructure, social media influence on prices, or on-chain behavioral patterns.

Practical Use Cases

Building a Wallet-Tracking Alert Bot

Subscribe to the fills WebSocket stream for a set of whale wallets and trigger alerts when significant trades execute. The fill data includes coin, size, price, and direction — everything needed for actionable notifications. Push alerts to Telegram, Discord, or any messaging platform when a tracked wallet opens a large position or realizes significant PnL.

Sentiment Analysis from KOL Tweets

Connect to the Twitter WebSocket with a minimum follower filter to focus on high-influence accounts. Feed tweets into a language model for sentiment classification, then aggregate scores by coin or pair. When sentiment shifts sharply — influential accounts turning bearish on a previously bullish asset — flag the divergence for review or trigger automated risk reduction.

Real-Time News-Driven Trading Signals

Each news item includes extracted trading symbols, making it straightforward to match events to assets. An agent that detects a listing announcement and correlates it with open interest data could generate a signal within seconds of the news breaking.

Copy Trading Automation

Combine the fills stream with the Trading API to build copy trading logic that goes beyond simple mirroring. Track top-performing wallets, analyze their fill patterns in real time, and execute copy trades with your own risk management rules — position sizing, leverage caps, asset filters, and timing adjustments.

Multi-Source Alpha Research

Merge the Twitter feed, news feed, and wallet fills data into a unified event stream. When a prominent KOL tweets about a specific asset, your system can simultaneously check whether tracked wallets are accumulating positions, whether open interest is rising, and whether the news feed contains corroborating information. This kind of multi-source convergence analysis is exactly what AI agents excel at.

Getting Started

The Agent API is available now at the dedicated Agent API page on HyperX. The page itself serves as both documentation and a live demonstration — when you visit, it automatically connects to the Twitter WebSocket feed and shows real-time data streaming in. You can see the data format, observe the update frequency, and verify connectivity before writing a single line of code.

The Twitter feed and news feed are completely free — no API key, no authentication, no rate limits. For the Data API endpoints and Trading API, generate an API key through the HyperX settings page. The free tier provides a generous weight-based rate limit budget suitable for development and low-frequency production use.

The page includes a complete, interactive endpoint catalog for both APIs with collapsible categories, plus five AI Skill cards — self-contained specifications covering the Twitter Feed, News Feed, Fills Stream, Data API, and Trading API. Copy any card directly into your agent's configuration for instant integration.

What Is Next

The Agent API as it launches today is the foundation of our programmatic access layer. We are actively expanding the available data streams, adding more granular filtering options, and building higher-level analytical endpoints that combine multiple sources into pre-computed signals.

The next generation of crypto market participants will not all be humans staring at charts. Many will be AI agents, automated systems, and hybrid workflows where humans set strategy and machines execute. The Agent API page is live now — visit it, watch the real-time data flow, copy a skill card, and start building.

Frequently Asked Questions

Is the Twitter feed really free? Yes. The Twitter / KOL feed and the News feed are both completely free with no authentication required. Connect to the WebSocket or call the REST endpoint — no API key needed, no rate limits applied.

What programming languages are supported? The Agent API uses standard REST and WebSocket protocols, so it works with any language that supports HTTP requests and WebSocket connections. The skill cards include examples in both Python and JavaScript, but the API is language-agnostic.

How do I get an API key for the Data API? Navigate to the HyperX settings page and generate your API key. The key is passed as a header with your requests. Free tier rate limits apply by default.

Can AI agents use the Trading API to execute real trades? Yes. The Trading API provides full programmatic access to order execution on Hyperliquid through HyperX's infrastructure. You will need a trading token, which is managed separately from the data API key for security isolation.

What is the latency of the WebSocket streams? WebSocket streams deliver data with sub-second latency from the source event. Twitter and news items appear within seconds of the original post or publication. Fills data streams as soon as the trade is confirmed on Hyperliquid.