// API & MCP

Trading Bot API & MCP Server

One API to create strategies, run backtests, optimize parameters and deploy live bots. MCP-native, so any AI agent framework can use Runbot as a trading tool in minutes.

See the AI Agent

REST API

Strategy CRUD, backtesting and optimization endpoints for building Runbot into your own product.

MCP Server

Expose Runbot's strategy, backtest and deployment tools to any AI agent that speaks the Model Context Protocol.

Webhooks

Real-time bot execution events and alerts pushed to your systems.

Docs & SDKs

Full documentation and SDKs for fast integration.

What the API lets you do

  • Create, read, update and delete trading strategies.
  • Run backtests on historical data and retrieve performance metrics.
  • Launch AI optimizations to tune strategy parameters.
  • Deploy strategies as live bots that run 24/7 in the cloud.
  • Receive execution events and alerts through webhooks.

An MCP server for trading

The Model Context Protocol (MCP) is an open standard that lets AI assistants and agents call external tools. Runbot's MCP server exposes strategy creation, backtesting, optimization and deployment as tools, so an AI agent can go from a user's request to a running bot on its own.

runbot-mcp-server
// AI Agent creates and deploys a strategy
const strategy = await runbot.createStrategy({
  name: "BTC RSI Scalper",
  market: "BTC/USDT",
  timeframe: "5m",
  indicators: ["RSI", "Orderflow"],
  stopLoss: 0.02,
});

const backtest = await runbot.backtest(
  strategy.id, { period: "2y" }
);

await runbot.deploy(strategy.id);
// Bot is live on the cloud 24/7

The snippet above is illustrative. Refer to the Runbot documentation for the current endpoints and SDK methods.

Use cases

  • Add AI trading features to an existing app without running your own backtesting or execution infrastructure.
  • Give an AI assistant or agent the ability to design, test and deploy strategies for its users.
  • Automate strategy research pipelines: generate variants, backtest them and keep the best performers.
  • Power a fully branded front end on top of Runbot with the white-label platform.

Frequently Asked Questions

MCP (Model Context Protocol) is an open standard that lets AI agents call external tools. With Runbot's MCP server, an AI agent can create, backtest, optimize and deploy trading strategies without custom glue code.

The Runbot documentation is available at docs.runbot.io.

Yes. Bots deployed through the API execute through exchange API keys or webhooks. Runbot never holds user funds.

Launch Your AI Trading Platform

Tell us about your project. Most partners go live within 1-2 weeks, and we handle the integration end to end.

How White-Label Works