QUANTCONNECT REMOTE MCP SERVER

Connect AI Agents to QuantConnect

A hosted Model Context Protocol server that lets your AI design, backtest, optimize, and live trade quantitative strategies on QuantConnect. Nothing to install, build, or run: point your MCP client at our endpoint, authorize your organization, and the tools are available immediately.

Getting Started

The remote server speaks the standard MCP HTTP transport, so any MCP-capable client can use it. To connect:

Add a new remote MCP server in your client and enter the endpoint https://www.quantconnect.com/api/v2/mcp. The configuration shown here is the format used by Copilot in Local Platform.
When the client prompts you to authenticate, open the authorization page in the browser that is logged in to your QuantConnect account.
In the Authorize MCP Server form, select the organization you want to connect to and click Authorize.
Open a project and start prompting. Try "Read the open project." to confirm the tools are connected.

Step-by-step guides are available for Claude Code, Copilot, Cursor, and Devin Desktop. The remote server requires a paid agent node (A1-1, A4-8, or A16-32) in your organization.

MCP SERVER CONFIG

{
  "servers": {
    "qc-mcp": {
      "url": "https://www.quantconnect.com/api/v2/mcp",
      "type": "http"
    }
  }
}

Connect Your AI Agent

Built for Our Quantitative Infrastructure

Designed with the needs of professional quant teams in mind, the Remote MCP Server gives your AI the same cloud infrastructure, data, and execution our platform runs on.

Projects and Files

Create projects from templates, read and edit files, and compile after every change to catch errors early.

Datasets and Skills

Browse dataset details and code examples, check preinstalled libraries, and load QuantConnect skills for LLM context.

Research Notebooks

Create, edit, and execute Jupyter cells in the QuantConnect Research environment, and read the results back.

Backtesting

Autonomously backtest ideas, search logs, review orders, and analyze results. Completed runs are delivered to your AI.

Parameter Optimization

Have your AI abstract parameters and test project sensitivity across combinations to protect from overfitting.

Live Trading

Deploy, monitor holdings and logs, stop, or liquidate live algorithms, with email, SMS, and Telegram notifications.

Explore the Full Tool Set

QC MCP Enables Powerful New LLM Workflows

The QuantConnect Remote MCP Server empowers AI to act like a trusted team member, able to execute tasks reliably on your behalf. Try some examples below to complement your research.

"Are there interesting new strategies being discussed online?"

Search research journals and QuantConnect discussions for recent interesting posts to create your next strategy.

"Review my live algorithms. Are there any we should stop trading?"

Fetch live algorithm results and review their performance, and perform analysis to see if any are outside of expected tolerance.

"Add Option hedging to my strategy for outlier events."

Analyze your project code and update it to purchase downside insurance. It will also run backtests to confirm reduced drawdown.

"Debug the last runtime error and fix the project code."

The MCP will fetch the last backtest, read its error message, modify the project to fix the issue, and re-run your backtest to confirm success.

Add MCP Agents For Your Team

Built to Empower Quants

A hosted, secure, and well-supported foundation for building AI enabled quant systems with confidence.

45+ Available Tools

A complete implementation of the QuantConnect API, plus AI-coding tools crafted by QuantConnect to reduce bugs in your LLM code.

Available Tools

Nothing to Install

Fully hosted by QuantConnect. No containers, API keys, or local processes to manage, and every client gets the latest tools automatically.

Supported Clients

Authorized Per Organization

Each client is authorized through your QuantConnect login and receives an access token scoped to your account and the organization you select.

Learn More

SIGN IN