Webull Cloud MCP Lets AI Prepare Trades Without Code Or…

Webull has broadened access to its cloud-hosted Model Context Protocol (MCP) server, allowing eligible U.S. customers to use compatible AI assistants to research markets, view authorized account data and prepare orders. The brokerage says users do not need to install software or manage API keys, and every order must still be confirmed separately by the customer before it is sent to the market.

The move shifts Webull's MCP offering from information access toward account action. A customer can ask about a company or portfolio and, within the same conversation, have an assistant assemble a trade instruction. Webull keeps a confirmation boundary between the language model and execution. The company previously offered a local MCP server that required users to install software, generate credentials and manage access keys. Cloud MCP moves that server into Webull's infrastructure and lets customers connect an existing account through a supported AI platform. That lowers the setup burden, potentially expanding the audience beyond developers and quantitative traders to retail investors who select the account and capabilities they want to authorize.

Cloud MCP can expose real-time and historical data for stocks, ETFs, options, futures, crypto and event contracts. It can also retrieve company fundamentals and filings, run screeners, manage watchlists, and show account balances, positions and order history where permission is granted. Because market data is relatively low risk while holdings and order history are sensitive, authorization design is critical. Webull's release does not detail session duration, audit logs, data retention by third-party assistants, or whether read-only account access and order preparation have separate permissions.

Anthony Denier, Group President and U.S. CEO of Webull, said the rollout makes it easier for investors to move from research to action through AI platforms they already use while keeping them in control of every trade. The control is the separate confirmation step: an assistant may prepare a buy or sell instruction, but the customer must approve it through Webull before submission. The design is similar to Leverate's MCP for Traders, which also requires explicit approval, and more constrained than systems that allow assistants to manage live orders and risk settings after connection. A confirmation screen should restate the instrument, side, quantity, order type, limit or stop price, time in force, estimated cost and material account effect so users can catch an imprecise prompt that became the wrong instruction.

MCP gives an AI application a standardized way to discover and call external tools. Webull provides the market and account functions, while the compatible assistant supplies the conversational interface. That separation may help Webull reach users across multiple AI products without building proprietary chat systems, but it also creates shared responsibility for prompt interpretation, conversation storage and user experience. Trading Central has used MCP to expose licensed research without connecting to account action; Webull's implementation goes further by combining information, private account data and order preparation.

Cloud MCP is available to eligible U.S. users through Claude, Perplexity and Grok, while ChatGPT access remains pending platform approval. Webull has also been adding AI across its platform, including AI-generated analysis in its research engine and a paper-trading sandbox for AI and automated strategies. The product's success will depend on whether permissions and the confirmation screen make the boundary real, and whether users can trace answers back to current market data, filings and their actual account information.