# QuantDinger MCP server

> Lets an AI client research markets, build and backtest Python trading strategies, and run paper or live multi-asset execution.

- Repository: https://github.com/OpenByteInc/QuantDinger
- Page: https://hostedmcp.org/servers/openbyteinc/quantdinger
- Transport: —
- Stars: 13734
- License: Apache-2.0

## FAQ

### Does QuantDinger have an MCP integration?

Yes, MCP is one of the supported access paths alongside the web UI, Human API, and Agent Gateway.

### What can I do with QuantDinger's factor analysis?

Factor Lab benchmarks up to 40 registered factors on a point-in-time universe panel using a time-ordered holdout, out-of-sample IC retention, and false-discovery control. The workflow is available from the web UI, Human API, Agent Gateway, and MCP.

### Can QuantDinger place real trades?

Yes, it can submit real orders when live trading is explicitly enabled, but paper trading is recommended to start. It does not provide investment advice.

### Are signal-only virtual accounts connected to a broker?

No, signal-only virtual accounts convert notification-mode signals into internal virtual orders, fills, positions, and PnL without connecting to a broker or submitting live orders.

### Is QuantDinger self-hosted?

Yes, it is designed as a local-first, self-hosted stack that keeps market data, strategy code, and credentials under the operator's control.