A quiet but significant revolution is taking place in the world of financial analytics. I am closely watching how AI tools are beginning to displace traditional, costly platforms, and the latest example is striking proof of this. An analyst under the pseudonym CyrilXBT demonstrated how, with just a single command, Claude Code can be connected to live market data covering more than 17,000 stocks and cryptocurrencies, as well as earnings reports, balance sheets, and cash flow statements. In essence, this replaces the capabilities of a Bloomberg terminal, which costs up to $24,000 per year.

Three steps to live data

The connection scheme described by CyrilXBT is striking in its simplicity. Through the MCP (Model Context Protocol), an external financial data service is integrated. The entire process takes about 60 seconds and consists of three stages.

The first step is running a command inside Claude Code to add the financial-datasets service via HTTP transport. The second step is authorization via OAuth in the browser. The third is natural language queries. For example, you can ask about Apple's current price-to-earnings ratio, request Tesla's quarterly reports for the last four periods, or find out Bitcoin's price dynamics over a year.

Previously, such analysis required either a Bloomberg terminal, a complex API, or hours of manual searching across various sources. Now, all of this is available in a single interface where AI not only provides figures but can also interpret them.

A quantum leap for analysts

A sustainable advantage will go to those who learn to combine Claude Code's analytical capabilities with access to real financial data. This concerns quants, analysts, and portfolio managers. This is not a one-time advantage—it accumulates daily, like a research edge that grows with every use of the combination.

Each step on its own is not complex, but value arises precisely from connecting an ordinary tool to a stream of up-to-date market data. Here, Claude Code acts as a shell through which the user accesses a third-party data provider without manual information gathering. This is an open standard, and its potential is enormous.

Scaling: thousands of agents overnight

Another aspect also deserves attention, which CyrilXBT revealed in his breakdown of the workflow of Boris Cherny, one of the creators of Claude Code. Cherny uses a smartphone as his primary interface instead of a laptop and maintains between five and ten active sessions simultaneously. Each of these can spawn auxiliary agents—sometimes several hundred, sometimes several thousand overnight for larger tasks.

Dozens of cycles run continuously in the background, with some of them preserved on the server side even when the laptop is closed. However, such a scheme requires strict conditions: a full instruction-sending interface, isolated parallel execution, verification without human involvement, and strict limits on time, money, and volume. This is what turns overnight agent work from a risk into a manageable task.

My conclusion: integrating AI with live financial data is not just convenience but a paradigm shift. Traditional terminals and expensive subscriptions are losing their exclusivity. The question is not who will first implement such combinations, but how quickly the market will adapt to a new reality where the speed of analysis becomes the main competitive advantage.