From spreadsheets to live market data, Anthropic’s LLM is becoming the default copilot for finance professionals.
Anthropic is doubling down on its lead in financial AI with a wave of updates that make Claude the most integrated LLM in financial services — spanning Excel modeling, real-time data, and automated analysis.
The company’s latest release introduces Claude for Excel, new connectors to leading financial data providers, and a suite of pre-built Agent Skills for core analytical workflows such as discounted cash flow modeling, due diligence, and earnings analysis. Together, they mark a decisive step toward embedding Claude directly into the daily tools and data pipelines of bankers, analysts, and portfolio managers.
Claude for Excel: A new analyst in the spreadsheet
Now in beta for Enterprise, Teams, and Max users, Claude for Excel brings the model directly into Microsoft Excel via a sidebar that can read, analyze, and modify workbooks in real time.
Claude can:
Explain how a spreadsheet works and highlight dependencies, Debug and fix formulas, Populate templates with new data and assumptions, Or build complete financial models from scratch.
Crucially, every action is transparent and traceable — Claude shows which cells it changed and why, turning Excel into a live, auditable AI workspace.
This integration builds on Anthropic’s broader Microsoft partnership: Claude can already interact with files, emails, and Teams conversations through Microsoft 365 and Copilot Studio. The Excel add-in simply brings that ecosystem closer to the front lines of financial analysis.
Connecting to real-time market intelligence
Anthropic is also expanding Claude’s connectors, giving the model governed access to live financial data across multiple sources.
New integrations include:
LSEG for equities, fixed income, FX, and macro indicators, Moody’s for credit ratings, ownership data, and company financials, Aiera for real-time earnings call transcripts and investor events, Third Bridge and Chronograph for expert interviews and private-equity performance data, MT Newswires for multi-asset market coverage, Egnyte for secure search across internal investment documents and data rooms.
These connectors transform Claude from a static reasoning engine into a real-time financial analyst, capable of blending internal and external data for valuation, research, or compliance tasks — without breaching governance constraints.
Pre-built Agent Skills for finance
To accelerate adoption, Anthropic has also released six new Agent Skills designed specifically for financial use cases. These modular “AI templates” automate key workflows:
Comparable company analysis with refreshable valuation multiples Discounted cash flow (DCF) models with scenario toggles and sensitivities Due-diligence data packs that extract structured insights from data rooms Company teasers and pitch-book profiles Earnings analysis from transcripts and financial reports Initiating coverage reports combining industry overviews and valuation frameworks
These skills can run across Claude.ai, Claude Code, and API environments, giving teams flexibility to deploy AI across research, corporate finance, and investment processes.
The emerging LLM standard for finance
Claude’s trajectory in financial services is clear: deep vertical integration rather than generic capability. By fusing Sonnet 4.5’s top-ranked performance on the Finance Agent benchmark (55.3% accuracy per Vals AI) with the workflows analysts actually use — Excel, FactSet, LSEG, Moody’s, Morningstar — Anthropic is positioning Claude as the go-to institutional AI layer for the financial sector.
Banks, asset managers, and insurers are already using Claude across front, middle, and back office: client engagement, underwriting, risk, compliance, and even code modernization.
The new suite of financial features pushes that further — turning Claude from a text model into an end-to-end financial workbench.
As other models chase general intelligence, Anthropic’s focus on governed, auditable, and domain-specific AI is quietly making Claude the default choice for serious financial work.
