Kensho LLM-ready API expands MCP server access, dataset support, integrations, and more
The Kensho LLM-ready API now features remote MCP server support, making S&P Global data a native connector in Claude and any MCP-compatible GenAI tool.
Last November, we launched the Kensho LLM-ready API — a purpose-built data retrieval solution designed to integrate seamlessly with large language models (LLMs) like GPT, Gemini, and Claude. The goal was clear: make S&P Global’s trusted tabular financial data accessible via natural language query wherever our customers need it.
Over the past year, the Kensho team has made significant progress rolling out new capabilities and adding support for a range of additional datasets. We’ve also learned a great deal from our customers and partners, whose feedback has helped guide our roadmap and momentum as we enable new AI use cases across financial services and look toward more complex agentic workflows.
Meeting our customers where they are also means enhancing the ease of integration. That’s why we launched a Model Context Protocol (MCP) server for the LLM-ready API, committing to this open standard and enabling developers to more easily build secure connections from their AI agents and LLMs to S&P Global data sources.
This server is the foundation of S&P Global’s strategic collaboration with Anthropic, giving our customers access to S&P Global’s trusted data and insights not only in Claude but any MCP-compatible GenAI application. Today we’re announcing remote MCP server support and an S&P Global Connector for Claude to simplify this process even further.
Streamlining access with remote MCP server connection
Starting today, it’s even faster and easier to install Kensho’s MCP server and access S&P Global data. With remote MCP server support for the Kensho LLM-ready API, you can now tap into S&P Global data more seamlessly in Claude using the new S&P Global Connector in Anthropic’s MCP Directory, or leverage remote server access via any MCP-enabled GenAI application.
For Claude users accessing S&P Global data through Anthropic’s Claude Pro subscription or Financial Analysis solution, you can set up S&P Global data access through the Kensho LLM-ready API in just a few clicks. Open Claude, click “add connectors,” select the S&P Global tile, click to connect, and log in to your authorized account.
The new S&P Global Claude Connector tile.
The new S&P Global listing in Anthropic’s MCP Partner Directory.
Kensho’s MCP tools are now available from all of Claude’s supported platforms including web, desktop, and mobile app. For additional details, check out this Anthropic article on using S&P Global data for financial analysis in Claude. More broadly, the power of our MCP server is that it enables customers to connect the LLM-ready API with any MCP-compatible system.
Expanding S&P Global dataset support
The Kensho LLM-ready API helps engineering, product, and business teams across financial services save time and resources by making company information, financial statements, historical market data, global securities data, and more just a question away.
At release last year, the API supported Financial Statements, Financial Line Items, Market Prices, and Price Charts from the S&P Capital IQ Financials and Market Data offerings. We viewed these as foundational to finance-specific GenAI applications. In 2025 we’ve expanded our coverage of those datasets to include segment-specific financials, market capitalization, total enterprise value (TEV), and more.
To support even more workflows, we added datasets such as Company Intelligence, Competitors, M&A Transactions, and Earnings Call Transcripts. These additions allow users to access data relevant to both private markets and unstructured text.
Support for private markets data unlocks entirely new capabilities for analysts and investors. By incorporating Company Intelligence and M&A transactions (with Private Company Financials coming soon), the LLM-ready API now serves a class of users previously underserved by GenAI tooling. Customers operating in sectors where disclosure is limited can now surface deeper insights, conduct diligence more efficiently, and explore markets that were previously difficult to map using traditional tools. This expansion empowers teams to move faster and make better-informed decisions in a domain that has long depended on manual workflows and siloed data.
Adding earnings call transcripts marked another major milestone. Unlike vector-based retrieval methods, which often return short excerpts or snippets, the LLM-ready API returns the full transcripts users request. This gives LLMs the ability to analyze complete documents, enabling quote extraction, summarization, sentiment analysis, trend analysis, and deeper document inspection in a single response. By maintaining the integrity and context of full documents, the API allows users to derive more meaningful insights from unstructured data.
Learning from customers as we build for an agentic future
All of this innovation comes on the heels of strong customer adoption, from large investment banks to private family offices. One of the most common use cases is completing public comparables, or comps.
Customers use natural language to retrieve comparable public company data and then apply GenAI to analyze those results. Many users highlight the consistency of our deterministic responses, which makes the API suitable even in audited environments. This customer feedback has been integral to our roadmap, and we’ll soon be enabling support for the Capital IQ Consensus Estimates dataset to further enhance this use case.
Looking ahead, we’re expanding the LLM-ready API’s capabilities to support emerging trends in agent-based workflows. As customers build multi-agent systems, we’re working to make even more data available to them. The team is currently developing an agent for the API that will integrate with Kensho’s Grounding Agent (currently in alpha), which uses a routing agent to direct queries to the most appropriate underlying tool. We look forward to sharing more updates on this front in the future.
If you’d like to test the LLM-ready API, reach out to commercial@kensho.com.