Introduction
The financial data layer for AI agents — first-party crypto market data in real time, multi-market coverage in one unified model, one API and one MCP, structured results ready for AI.
GateData.ai delivers a clean, stable, predictable financial data API — one REST API and one native MCP Server so AI agents can read real-time, structured, machine-ready market data directly.
Core crypto market data comes from Gate’s own matching engine: prices, order books, trades, and liquidity are collected first-party and exposed via API within seconds of being observed. Beyond crypto, GateData.ai also covers US and Hong Kong equities plus macro data — 4,700+ assets queryable.
Everything sits on one unified object model: entities, assets, instruments, venues, protocols, chains, and wallets share a consistent schema. An agent can jump from a Gate trading pair to a project entity, a US stock, or a prediction event — without switching data models or rewriting integration logic per market.
What You Can Access
Data is organized into domains, not a loose set of endpoints. Browse the full Data Catalog to see every domain, its capabilities, and example calls.
- Market Data & Liquidity — Query crypto market data and liquidity, including live prices and klines, order books and trades, plus derivatives and liquidations; US and Hong Kong equities include snapshots and klinesLive
- Prediction Markets — Tradable events, rankings, event signals, and current order books across major prediction venues.Live
- Asset & Entity Reference — Reference data for assets, tokens, companies, projects, protocols, chains, venues, and wallets, unified under one object model.Live
- Company Fundamentals — Company profiles, financial statements, metrics, segments, and corporate actions.Live
- Estimates & Earnings — Query earnings timing and reported results for US-listed companies, with consensus estimates for EPS and revenueLive
- Network & Protocol Activity — Coming soon. Planned: on-chain and DeFi metrics (TVL, fees, stablecoins, bridges, yield pools, and exchange reserves).Coming soon
- Ownership, Flows & Wallet Intelligence — Coming soon. Planned: token holders, address activity, wallet labels, and fund flows.Coming soon
- Events, News, Sentiment & Signals — Structured news, events, social sentiment, and social posts.Live
- Risk, Compliance & Disclosures — Coming soon. Planned: contract security, address risk screening, and disclosures.Coming soon
Built For
- AI agents and LLM / MCP pipelines
- Quantitative research and trading platforms
- Portfolio and risk management systems
- Event-driven and signal-based strategies
- Internal data layers for crypto and fintech products
Design Principles
AI-native delivery
Public domains are available via REST and MCP so agents can call them without glue code. Network, Wallets, and Risk are coming soon.
Unified object model
Assets, instruments, venues, protocols, chains, and wallets share one consistent schema across all domains.
First-party market data
Core crypto market data comes straight from the Gate exchange, not relayed through third parties.
Machine-first JSON
Consistent field naming, normalized units, and English output across all endpoints, optimized for programmatic consumption.
Real-time
Crypto market data is served with low latency; time semantics are explicit on every response. On-chain domains are coming soon.
Documentation
GateData.ai developer docs — authenticate, call REST, or connect via MCP.
Sign in, open My Account, create an API key, and make your first call.
5-minute setup →Public and upcoming domains — capabilities, REST paths, and MCP tools.
Browse data catalog →REST endpoints, parameters, response fields, and example calls.
Open API Reference →Install and configure MCP in Cursor, Claude, Codex, and other clients.
MCP docs →Entity, asset, instrument, listing, venue, protocol, wallet, and signal objects.
View data model →Use Case Templates and how-tos — from first API call to runnable crypto and research scenarios.
View templates →Machine-readable API specification for SDK generation and client tooling.
OpenAPI spec →