🎉 Our Chrome Extension is here! Get live market prices right in your browser.Install Now
Everything you need to build production-grade market data applications.
Persistent connection for price, candles, order-flow imbalance, market-wide, and multi-timeframe streams.
Standard JSON responses for current price, OHLC candles, and market snapshots.
Query H1 historical OHLC data — paginated, no current bar.
Native Model Context Protocol server with 34 quota-free tools for AI agents, analytics, and advanced market context.
Every access method delivers the same underlying data stream — choose the protocol that fits your stack and latency requirements.
Standard HTTP GET requests returning JSON. Retrieve the current price snapshot for any symbol, request a paginated slice of historical OHLC candles, or fetch pre-computed indicator values — RSI, MACD, Bollinger Bands, ATR, Stochastic — with a single authenticated call. Responses include timestamps in ISO 8601 UTC, bid/ask spread, volume, and OHLCV fields. Rate limits apply per-minute based on your subscription plan. Ideal for server-side rendering, scheduled jobs, mobile apps, and any environment where a persistent connection is impractical.
REST API reference →Persistent wss:// connections that push price and candle updates as they occur. Once subscribed, you receive a new message for every completed candle on your chosen timeframe — M1, M5, M15, H1, H4, D1 — plus live tick updates when price moves by the minimum threshold. A single connection can multiplex multiple symbols. Reconnect logic with exponential back-off is straightforward to implement. Available on Plus plans and above. Recommended for live dashboards, alerting bots, and charting front-ends that need sub-second data.
WebSocket streaming docs →A native Model Context Protocol server exposing 34 market data tools that any tool-calling LLM can invoke autonomously. Connect Claude, GPT-4o, Gemini, or a custom agent and it gains the ability to fetch live prices, retrieve historical OHLC candles, compute technical indicators, and identify support/resistance levels — all as part of its reasoning chain. No manual prompt engineering for data retrieval. Tool calls count against your plan limits. Available on Pro plans and above.
MCP server docs →