# WorldQuant BRAIN (unofficial)

> WorldQuant BRAIN (unofficial) is an MCP server for finance and payments. Backtest WorldQuant BRAIN alphas and browse data fields from an agent (unofficial). It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Platforms:** CLI, API
- **Website:** https://pypi.org/project/mcp-server-worldquantbrain/
- **Repository:** https://github.com/tukorama/mcp-server-worldquantbrain
- **MCP Registry name:** io.github.tukorama/mcp-server-worldquantbrain
- **Categories:** [finance and payments](https://ai-powered.com/categories/finance-and-payments), [analytics and data](https://ai-powered.com/categories/analytics-and-data)
- **Listing updated:** 2026-11-14

## About

This unofficial server wraps the WorldQuant BRAIN API so an agent can look up data fields and operators, run Fast Expression alpha simulations and read results. It is not affiliated with WorldQuant, and a BRAIN account is required.

The design reacts to how agents tend to misuse the platform. Simulations take minutes, so a timeout returns PENDING with an id and get_simulation_status picks it up without resubmitting. A local ledger remembers every expression and settings pair, so a repeat returns the earlier alpha instead of spending metered quota. Large responses are trimmed: a PnL series becomes yearly figures and lists are capped.

Submission is switched off unless WQB_ALLOW_SUBMIT is true, and the bundled LangGraph integration pauses for human approval before submit_alpha. The same tools are available as LangChain tools with a prebuilt agent. It installs from PyPI.

## Features

- **Alpha backtesting:** Runs FASTEXPR alpha simulations on BRAIN.
- **Resumable simulations:** A timed-out wait returns PENDING and an id to poll.
- **Repeat protection:** A local ledger returns earlier results instead of spending quota again.
- **Trimmed results:** Large payloads are summarised, such as yearly PnL.
- **Gated submission:** submit_alpha appears only when WQB_ALLOW_SUBMIT is true.
- **LangGraph tools:** Same tools as LangChain tools plus a prebuilt agent.

## Good for

- Backtest a quant expression from a chat
- Browse available data fields and operators
- Avoid re-running identical simulations
- Build a LangGraph research agent with an approval step before submission

## Install

### Claude Code (Terminal)

```sh
claude mcp add mcp-server-worldquantbrain --env WQB_PASSWORD='<WQB_PASSWORD>' -- uvx mcp-server-worldquantbrain
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "mcp-server-worldquantbrain": {
      "command": "uvx",
      "args": [
        "mcp-server-worldquantbrain"
      ],
      "env": {
        "WQB_PASSWORD": "<WQB_PASSWORD>"
      }
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "mcp-server-worldquantbrain": {
      "command": "uvx",
      "args": [
        "mcp-server-worldquantbrain"
      ],
      "env": {
        "WQB_PASSWORD": "<WQB_PASSWORD>"
      }
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "mcp-server-worldquantbrain": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "mcp-server-worldquantbrain"
      ],
      "env": {
        "WQB_PASSWORD": "<WQB_PASSWORD>"
      }
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "mcp-server-worldquantbrain": {
      "command": "uvx",
      "args": [
        "mcp-server-worldquantbrain"
      ],
      "env": {
        "WQB_PASSWORD": "<WQB_PASSWORD>"
      }
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "mcp-server-worldquantbrain": {
      "command": "uvx",
      "args": [
        "mcp-server-worldquantbrain"
      ],
      "env": {
        "WQB_PASSWORD": "<WQB_PASSWORD>"
      }
    }
  }
}
```

## Tools

- `get_simulation_status`: Resumes a pending simulation by id.
- `submit_alpha`: Submits an alpha; only available when WQB_ALLOW_SUBMIT is true.

## FAQ

### Is it official?

No. It is unofficial and not affiliated with WorldQuant.

### How do I log in?

Run mcp-server-worldquantbrain auth, or set WQB_EMAIL and WQB_PASSWORD, or point WQB_CREDENTIALS_FILE at a JSON file.

### Can the agent submit alphas?

Only if WQB_ALLOW_SUBMIT=true; the LangGraph agent then asks for human approval.

### What if a simulation takes too long?

It returns PENDING with an id; use get_simulation_status to resume.

### How do I add WorldQuant BRAIN (unofficial) to Claude Code?

Run this command in your terminal: claude mcp add mcp-server-worldquantbrain --env WQB_PASSWORD='<WQB_PASSWORD>' -- uvx mcp-server-worldquantbrain

### What are the best alternatives to WorldQuant BRAIN (unofficial)?

The closest alternatives to WorldQuant BRAIN (unofficial) are QuantWright, QuantWorlds, quant-swarm, QuantBloc and QuantGPT.

## Where this comes from

- Description: written from the project's README and the project's website.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [QuantWright](https://ai-powered.com/mcp/quantwright): QuantWright is a paid MCP server for finance and payments.
- [QuantWorlds](https://ai-powered.com/mcp/quantworlds): QuantWorlds is an MCP server for finance and payments.
- [quant-swarm](https://ai-powered.com/mcp/quant-swarm): quant-swarm is a freemium MCP server for finance and payments.
- [QuantBloc](https://ai-powered.com/mcp/quantbloc): QuantBloc is a paid MCP server for finance and payments.
- [QuantGPT](https://ai-powered.com/mcp/quantgpt): QuantGPT is an MCP server for finance and payments.
- [FinzData](https://ai-powered.com/mcp/finzdata): FinzData is a freemium MCP server for finance and payments.
- [QuanterLab](https://ai-powered.com/mcp/quanterlab): QuanterLab is an MCP server for finance and payments.
- [Canli Research](https://ai-powered.com/mcp/canli-research): Canli Research is an open-source MCP server for finance and payments.

---
Source: https://ai-powered.com/mcp/worldquant-brain-unofficial
