SEC MCP Server · Financial Intelligence · Limited Beta

The SEC MCP Server.
Ask your AI anything.

The SEC MCP Server connects AI agents directly to live EDGAR filings — income statements, balance sheets, cash flow, multi-company comparisons — all queryable in plain language. No spreadsheets. No data exports. No wrangling.

↓   scroll
What makes it different

Financial intelligence no other service delivers

Every financial data API gives you raw numbers. The SEC MCP Server gives your AI the ability to reason about those numbers — structured, sourced directly from EDGAR filings, and delivered as context your AI can use, not a spreadsheet you have to interpret yourself.

📡
Live EDGAR Data

Sourced directly from SEC EDGAR — the same filings institutions use. Income statements, balance sheets, and cash flow for every public company, updated as filings are submitted.

🗣️
Natural Language Queries

Your AI asks. The server answers. No query language, no formulas, no data schema to memorize. Just ask what you want to know and get structured financial data back instantly.

⚖️
Multi-Company Comparison

Compare any two or more public companies across any financial metric in a single query. Something no spreadsheet, no dashboard, and no other AI tool can do in one step.

🔒
Private and Isolated

Your queries are private. No shared infrastructure, no logging of your research. What you ask stays between you and the server.

Output Examples

What your AI can do with this server

TradingView, Trade Ideas, Bloomberg — every platform on the market works with pre-structured numerical data or price patterns. None of them can read 10-K narrative sections, parse proxy filings, or screen by derived fundamentals you define on the fly. Here's what that unlocks.

01 Risk Factor Evolution — 10-K Text + Numbers Combined
Query

"Track how Apple described China risk in their 10-K risk factor sections from 2019 to 2024, and show how their actual China revenue percentage changed over the same period."

Server Response
Year 10-K Risk Language (excerpt) China Revenue % Signal
2019 "tariff and trade policy uncertainty may adversely affect demand" 17.8% Neutral
2021 "dependence on third-party manufacturers in China… supply chain concentration risk" 19.2% ↑ Growing
2023 "geopolitical tensions and increased regulatory scrutiny in China may restrict our ability to conduct business" 17.5% ↓ Declining
2024 "intensifying competition from domestic Chinese smartphone manufacturers… market share pressure in Greater China" 15.1% ↓ Accelerating
AI synthesis: Apple's risk factor language tracked its actual China exposure with a 12–18 month lead time. The shift from "tariff uncertainty" language in 2019 to explicit "domestic competitor" language in 2024 preceded a 3-point revenue share decline. No charting platform, screener, or financial terminal reads 10-K narrative text — this query class is only possible via the SEC MCP Server.
02 Quality Earnings Screen — Sector-Wide Derived Metrics
Query

"Find S&P 500 software companies where free cash flow exceeded net income by more than 20% for three consecutive years AND gross margins improved each year. Rank by FCF premium."

Server Response — Top Results
Company Avg FCF / NI Premium Gross Margin Trend 3yr FCF CAGR Quality Score
Visa (V) +48% ↑ 79% → 81% +12.4% ★★★★★
Microsoft (MSFT) +41% ↑ 68% → 72% +18.7% ★★★★★
Intuit (INTU) +35% ↑ 79% → 83% +9.8% ★★★★☆
Automatic Data (ADP) +31% ↑ 45% → 48% +8.2% ★★★★☆
AI synthesis: FCF consistently exceeding GAAP net income indicates low accrual-based earnings manipulation — a quality signal institutional analysts pay for. TradingView, Trade Ideas, and every scanner on the market screen for price and volume patterns. This screen runs multi-condition fundamental quality filters across 500 companies from raw EDGAR data in a single query. That capability does not exist anywhere else.
03 CEO Pay vs. Performance — Proxy Filing Intelligence
Query

"For the 10 largest S&P 500 companies, compare CEO total compensation growth to total shareholder return over the last 5 years. Which executives are most disconnected from performance?"

Server Response — Sourced from DEF 14A Proxy Filings
CEO / Company 5yr Comp Growth 5yr TSR Pay-Performance Gap
Jensen Huang / NVDA +312% +1,840% Aligned ↑
Tim Cook / AAPL +18% +290% Aligned ↑
Andy Jassy / AMZN +220% +68% Misaligned ⚠
Sundar Pichai / GOOGL +195% +112% Mild gap ⚠
AI synthesis: This analysis requires parsing DEF 14A proxy statements — SEC filings that disclose executive compensation, equity awards, and pay-ratio data. Every major financial platform (Bloomberg, FactSet, Benzinga) works with trading data and earnings numbers. None of them index proxy filings for natural language queries. This is genuinely dark data for retail investors — until now.
04 Capital Efficiency Screening — Custom Metrics Across 7 Companies
Query

"Analyze capital efficiency trends across the Magnificent 7 tech companies over the last 5 years. For AAPL, MSFT, GOOGL, AMZN, META, NVDA, and TSLA: calculate the ratio of Free Cash Flow to R&D Expense for each year, show which companies generate the most cash per dollar of R&D investment, and identify any companies where capex is growing faster than revenue."

Server Response — Custom FCF/R&D Efficiency Analysis
Company FY2021 FCF/R&D FY2025 FCF/R&D Trend Efficiency Score
NVDA 1.48× 5.55× ↑ +275% ★★★★★
AAPL 4.25× 2.86× ↓ −33% ★★★★☆
MSFT 3.12× 2.94× ↓ −6% ★★★☆☆
META 2.18× 1.87× ↓ −14% ★★★☆☆
GOOGL 2.89× 2.41× ↓ −17% ★★★☆☆
TSLA 2.46× 0.94× ↓ −62% ★★☆☆☆
AMZN N/A* N/A* Data issue

*Amazon classifies R&D as "technology and infrastructure" expense rather than a separate R&D line item — server auto-flagged this accounting quirk.

AI synthesis: NVIDIA is the runaway winner — its FCF/R&D efficiency exploded from 1.48× to 5.55× as the AI GPU supercycle scaled revenue 13× while R&D only grew ~5×. Apple has the most consistent efficiency but is in slow decline as R&D spending accelerates faster than free cash flow growth. Tesla shows a collapsing efficiency profile (2.46× → 0.94×), with R&D spending now nearly equal to all free cash generated — a deteriorating margin story. Microsoft and Google both show positive operating leverage, but their massive AI infrastructure buildouts (capex growing 33% and 32.5% CAGR respectively vs. 14-17% revenue growth) represent significant capital deployment bets that either pay off in AI revenue or become write-down risks. This analysis required pulling 5 years of cash flow and income statement data across 7 companies, computing a custom ratio not available in any screener, and flagging Amazon's non-standard R&D classification — impossible on Bloomberg without manual Excel work, not possible at all on Yahoo Finance.
Platform Comparison

SEC MCP Server vs. The Competition

Bloomberg Terminal costs $25,000/year and requires Excel exports for custom analysis. Yahoo Finance is free but limited to single-company views. The SEC MCP Server delivers institutional-grade financial intelligence with zero manual work.

Capability
Beta · No Pricing Yet
$25,000/year
Free
Natural Language Queries Ask questions in plain English, get structured answers
YES Full natural language
NO Terminal commands
NO Manual navigation
Custom Ratio Calculations Define new metrics on the fly (FCF/R&D efficiency, operating leverage, etc.)
YES Any ratio, instantly
PARTIAL Export to Excel
NO Pre-set ratios only
Multi-Company Comparison Compare 7+ companies across any metric in one query
YES Unlimited, one query
PARTIAL Manual export + merge
NO One at a time
Derived Fundamental Screening Multi-condition filters (e.g., "FCF > NI by 20%+ for 3 years AND margins improving")
YES Complex conditions
PARTIAL Build in Excel
NO Not possible
AI Pattern Detection Auto-identify trends (NVDA's AI supercycle, TSLA's efficiency collapse)
YES Built-in synthesis
NO You do analysis
NO Raw data only
Accounting Quirk Awareness Flags when companies classify metrics differently (AMZN's R&D as "tech infrastructure")
YES Auto-flagged
NO Must know quirk
NO Not available
Time to Insight 7-company capital efficiency analysis with 5-year trends
YES 30 seconds
PARTIAL 2-3 hours
NO Not possible
SEC Filing Coverage 10-K, 10-Q financial data from EDGAR
YES 6,883 companies, 5.9M facts
YES Full EDGAR
PARTIAL Basic financials
Manual Work Required Spreadsheets, exports, formula building
NONE Zero manual work
HIGH Excel required
HIGH All manual
Query Volume Limits Number of analyses you can run
UNLIMITED No restrictions
LIMITED Per-seat license
LIMITED Rate limited
Limited Beta · No Pricing Yet

Request early access

The SEC MCP Server is in limited beta. We're onboarding a small group of analysts, investors, and founders who want financial intelligence built into their AI workflow — not bolted on. No pricing yet. Just early access.

Request Beta Access

Tell us how you'd use it. That's all we need.