Best Fundamentals APIs in 2026 (8 APIs Compared)

Search for the best fundamentals API and you will find a dozen vendors that look almost identical on the surface: each returns an income statement, a balance sheet, and a cash flow statement as clean JSON over a REST API. The real differences are underneath, in where the data comes from, whether a value can be traced back to the SEC filing it was parsed from, what actually sits on the free tier, and how the pricing is structured.

This is our roundup of the eight fundamentals APIs worth knowing in 2026. We build one of them, StockFit, and we rank it first, so here is the honest framing up front: every competitor claim below is checked against that vendor's own pricing and documentation as of 2026-07-14, we say plainly what StockFit does not do, and the one differentiator we lead with is a structural, verifiable fact rather than a superlative. If you just want the quick answer, jump to the comparison table.

A quick definition, because the category names blur together. A fundamentals API (also called a fundamental data API or financial statements API) serves company fundamentals: the audited financial statements, the derived financial ratios and valuation metrics, and usually earnings, dividends, and historical data going back years. That is different from a market data API, which serves prices, OHLCV data, and real-time quotes. Several vendors here do both. If you are new to the underlying analysis, our primer on fundamental stock analysis covers what these inputs are used for.

How to choose a fundamentals API

Before the vendor list, here is what actually separates one fundamentals API from another. These are the criteria the comparison scores against.

• Data source and traceability. Is the data parsed from SEC filings and XBRL, or aggregated and normalized from third-party sources? And can you trace any value back to the filing it came from? This is the single biggest quality axis, and the one most APIs are quiet about.
• As-reported vs normalized. Normalized data maps every filer onto one schema so a query works across thousands of companies. As-reported financial statements preserve the original XBRL concept names and values exactly as filed, including company-specific extension tags. Auditing needs as-reported; cross-company screening needs normalized. The best APIs give you both.
• Point-in-time correctness. For backtesting, every value has to be tagged with the date it became public, and the original figure preserved when a later filing restates it. Without point-in-time data you get look-ahead bias, which is covered in depth in our point-in-time data guide.
• Free tier. How many calls per day, and does it actually include fundamentals? Several APIs advertise a free tier that excludes financial statements entirely, or limits them to a handful of tickers.
• Coverage and historical depth. Quarterly as well as annual statements, and how many years of history, often gated by plan. Check the tier you are actually pricing.
• Statements, ratios, and metrics. Beyond the three statements, do you get financial ratios, valuation metrics, and key metrics computed for you, or do you derive them from raw line items? See /api/financials/key-metrics for the shape of a sector-aware metrics response.
• Developer experience. Clean structured JSON, a real REST API, sane rate limits, good documentation, and increasingly a Model Context Protocol (MCP) server so AI agents can call it. MCP is now common, so it is table stakes rather than a differentiator.
• Scope honesty. Some of these vendors bundle real-time market data, intraday and OHLCV data, technical indicators, options chains, economic indicators, and WebSocket streaming. A pure fundamentals API does not, and does not pretend to. Decide which half of that you need before you pay for the other half.

The best fundamentals APIs in 2026 at a glance

Eight APIs, scored on the axes that matter for SEC fundamentals. Prices are the cheapest plan that includes genuine financial statements, verified against each vendor's own pricing page on 2026-07-14. Prices drift, so re-check before you commit.

Notes: ¹ FMP's $22/mo is the annual-billed rate; its month-to-month Starter rate is not published. FMP's free plan covers prices, not fundamentals, and Starter is annual-period statements only; full quarterly plus deep history starts at Premium (about $59/mo billed annually). Alpha Vantage exposes full normalized statements on its free key but caps it at 25 requests per day. Finnhub's free tier includes SEC as-reported statements and key metrics, but the clean normalized statements are a paid add-on. Intrinio's free access is a non-production Dow-30 sandbox, and its Individual plan was recently revamped to $150/mo. Financial Datasets AI is free for five tickers and cites the source filing per statement. sec-api.io's free tier is a one-time 100-call trial.

1. StockFit, best for audit-grade, SEC-traceable fundamentals

StockFit parses company fundamentals directly from SEC EDGAR: raw XBRL out of the 10-K, 10-Q, and 20-F, run through the open-source Arelle engine, classified into fiscal periods, and served as both a normalized statement and the as-reported fact tree. What makes it different from every other API on this list is what rides along with each number.

Every value carries the SEC accession number and filing date of the filing it was parsed from, and when a later filing restates a figure the prior value is kept with its own accession number. That is the audit trail, and it is what lets you reconstruct exactly what was knowable on any past date. The chart below is six years of Apple revenue: every bar traces to a specific 10-K accession number, shown on hover.

In the response, that trail lives in a sources map keyed by accession number. Here is a real, trimmed response from /api/financials/income-statement for Apple's FY2025:

curl -H "Authorization: Bearer $KEY" \
  "https://api.stockfit.io/v1/api/financials/income-statement?\
symbol=AAPL&period=annual&limit=6"

[
  {
    "period":       "2025-09-26",
    "fiscalYear":   2025,
    "fiscalPeriod": "FY",
    "dateFiled":    "2025-10-31",
    "facts": {
      "revenue":         416161000000,
      "operatingIncome": 133050000000,
      "netIncome":       112010000000,
      "eps":             7.49,
      "epsDiluted":      7.46
    },
    "sources": {
      "0000320193-25-000079": {
        "type":      "10-K",
        "dateFiled": "2025-10-31",
        "amendment": false,
        "facts":     { "revenue": {}, "netIncome": {}, "epsDiluted": {} }
      }
    }
  }
]

That accession number, 0000320193-25-000079, is Apple's FY2025 10-K on EDGAR. You can open the filing and confirm the revenue figure yourself. When you want the raw filing instead of the curated shape, /api/financials/as-reported returns the original XBRL concept names including company-specific extensions, the same primitive our XBRL-to-JSON backtesting guide builds on. Financial ratios and valuation metrics come pre-computed from /api/financials/key-metrics, and the underlying /api/filings index is free.

2. Financial Modeling Prep, best for breadth at a low price

Financial Modeling Prep (FMP) is the broadest low-cost option. One subscription covers standardized financial statements, financial ratios, prices, market news, ETFs, 13F holdings, and earnings transcripts, with an official MCP server for AI agents. The data is extracted from SEC filings and standardized in-house, and there is a separate as-reported endpoint for the as-filed values.

The catch is where fundamentals sit in the pricing. The free plan (250 calls/day) is prices and reference data, not financial statements. The cheapest fundamentals tier, Starter, is annual-period statements only with up to five years of history, at about $22/mo billed annually. Full quarterly statements and 30 years of history start at Premium, about $59/mo billed annually. Coverage is genuinely broad, but the standardization is FMP's own, there is no per-value accession trail, and standardization accuracy across thousands of filers is a real consideration for financial-statement work. For the metrics side specifically, we compared the two directly in sector and industry metrics vs FMP.

3. EODHD, best all-in-one global bundle

EODHD (EOD Historical Data) is the all-in-one infrastructure play. Its Fundamentals Data Feed combines financial statements, valuation ratios, dividends, EPS history and estimates, insider transactions, and institutional holdings across more than 70 global exchanges and 120,000-plus tickers, all normalized into one schema, with an official MCP server. Historical data runs back to 1985 for major US companies.

Pricing is straightforward: the free tier is 20 calls per day and excludes fundamentals (a demo key exposes them for about six fixed tickers), and the Fundamentals Data Feed is $59.99/mo, or $49.99/mo billed annually. It is a strong choice when you want global fundamentals plus EOD prices under one bill. The trade-off is the same structural one: the data is third-party normalized from many sources, so it is convenient but not traceable to a specific SEC filing, and there is no point-in-time accession trail.

4. Alpha Vantage, best free tier to prototype with

Alpha Vantage is the default first stop for good reason: it is the one vendor here besides StockFit that exposes full normalized financial statements on its free key. Income statement, balance sheet, cash flow, a company overview with 50-plus ratios, earnings with estimates and surprises, dividends, and splits are all reachable free, with deep history (a live IBM pull returns 20 annual and 81 quarterly reports), and an official MCP server.

The constraint is throughput. The free tier is capped at 25 requests per day, and the paid tiers ($49.99/mo, or $41.58/mo billed annually, and up) buy only a higher rate limit, not more or better fundamentals. The statements are normalized to GAAP and IFRS taxonomies with no as-reported view and no accession trail, and there is no batch parameter, so pulling a large universe means one symbol per request. We put the two APIs side by side, endpoint for endpoint, in StockFit vs Alpha Vantage.

5. Finnhub, best for fundamentals plus context

Finnhub pairs company fundamentals with news, sentiment, and alternative data, which makes it attractive when you want contextual intelligence alongside the statements. Its free tier is generous on rate (60 calls per minute) and, unusually, includes SEC financials-as-reported and a large basic-financials block of key metrics and ratios.

The nuance is that the clean normalized income statement, balance sheet, and cash flow (30-plus years) are a paid product, the Fundamental Data add-on at $50/mo. So free gets you the raw as-reported values and metrics; the standardized statements cost money. The as-reported endpoint keys each report to a single filing-level accession number rather than per value, and there is no official Finnhub MCP server yet, only community ones. Good value if you want raw SEC as-reported data free and standardized statements cheap.

6. Intrinio, best for standardized and as-reported enterprise data

Intrinio is the closest competitor to StockFit on data source: fundamentals come directly from the SEC, XBRL-sourced with coverage going back about 15 years to the SEC's XBRL mandate, and offered in both standardized and as-reported form with 130-plus calculated metrics and statement notes. Turnaround is fast, roughly 10 to 30 minutes after a filing hits EDGAR, and there is an MCP server included with every plan.

Where it differs is packaging and price. There is no production free tier, only a non-production Dow-30 sandbox, and the entry plan, Individual, is $150/mo (recently revamped, so verify the current number). The standardized response carries a filing date and CIK but not an accession number per value, so tracing a figure to its source means a separate call to the Filings endpoints. It is a solid, institutional-grade option if the budget is there and you want both as-reported and standardized statements from one SEC-native provider.

7. Financial Datasets AI, best AI-agent-native option

Financial Datasets AI is built for LLM agents from the ground up. It ingests SEC filings and returns income statements, balance sheets, and cash flow statements in both a normalized schema and an as-reported line-item tree, plus segmented financials by business unit and geography, with an MCP server and clean structured JSON. Usefully, each returned statement carries a per-filing accession number and filing URL, so an agent can cite the source, which is a real differentiator on the AI angle.

Free access is limited to five tickers (AAPL, MSFT, NVDA, TSLA, GOOGL), which is fine for prototyping and no good for coverage. Paid entry is a $20 one-time credit pack (1,000 requests, one year of history); the full 30-plus years of quarterly and annual history is the Build plan at $200/mo. The citation is per filing rather than per value, and startup pricing changes often, so confirm the current plan structure. The strongest fit is an AI research agent that needs SEC-grounded statements it can cite.

8. sec-api.io, best SEC filings and XBRL specialist

sec-api.io is not a fundamentals product in the turnkey sense; it is a complete SEC EDGAR API, and its XBRL-to-JSON converter is how you get financial statements out of it. Point it at a 10-K or 10-Q and it returns the income statement, balance sheet, cash flow, comprehensive income, and equity statements as as-reported line items, with concept names normalized to the root US-GAAP taxonomy. Around that sits every EDGAR form type, full-text search, and real-time filing streams.

It returns no pre-computed financial ratios or blended fundamentals: you assemble the numbers you need from the raw XBRL yourself. The free tier is a one-time 100-call trial rather than a recurring allowance, and the entry plan (Personal and Startups) is $55/mo, or $49/mo billed annually. There is no MCP server. It is the right tool when you want raw, high-fidelity access to the whole of EDGAR rather than a curated fundamentals feed. We drew that line more fully in StockFit vs sec-api.io.

Which fundamentals API should you choose?

• You need traceable, point-in-time fundamentals for backtesting, screening, or auditing: StockFit. It is the only one that attaches an SEC accession number and amendment trail to every value, with the statements on the free tier.
• You want the widest cheap coverage across data types: Financial Modeling Prep, if you can accept in-house standardization and fundamentals starting around $22 to $59/mo.
• You need global fundamentals plus EOD prices in one bill: EODHD.
• You are prototyping and want free normalized statements today: Alpha Vantage, if 25 requests per day is enough.
• You want free SEC as-reported statements and cheap normalized ones: Finnhub.
• You are an enterprise wanting standardized and as-reported from one SEC-native vendor: Intrinio.
• You are building an AI research agent that must cite its source: Financial Datasets AI, or StockFit for a per-value citation.
• You want raw access to all of EDGAR and will build the fundamentals layer yourself: sec-api.io.

Two honorable mentions and one baseline. Twelve Data and Tiingo also pair fundamentals with market data at modest prices, and Polygon (now Massive) is market-data-led with some fundamentals attached. And the baseline everyone builds on is free: the official SEC EDGAR API serves the raw XBRL company facts at no cost, capped at 10 requests per second. If you are willing to do the period classification, quarter reconstruction, and multi-class handling yourself, it is the source under all of these. The paid APIs exist because that parsing is the hard part.

FAQ

What is the best fundamentals API in 2026?

It depends on what you need, but for accuracy and auditability StockFit is our pick: it parses fundamentals directly from SEC EDGAR XBRL, returns both as-reported and normalized financial statements, and ties every value to the SEC accession number and filing date of its source filing, with the amendment trail for point-in-time work, all on a free tier of 750 calls per day. Financial Modeling Prep and EODHD are strong low-cost all-in-one options, Alpha Vantage has the most generous free fundamentals for prototyping, Intrinio and Financial Datasets AI are SEC-sourced with as-reported statements, and sec-api.io is the specialist for raw XBRL-to-JSON. Match the API to whether you need traceability, breadth, or the lowest price.

What is the best free fundamentals API?

StockFit gives the most usable free fundamentals: 750 API calls per day with audit-grade income statements, balance sheets, cash flow statements, ratios, and the as-reported XBRL, no credit card. Alpha Vantage also exposes full normalized statements on its free key but caps it at 25 requests per day. Finnhub's free tier includes SEC as-reported statements and key metrics but gates the clean normalized statements behind a paid plan, and Financial Datasets AI's free access is limited to five tickers. Most other vendors (FMP, EODHD, sec-api.io, Intrinio) either exclude fundamentals from their free plan or offer only a short trial or non-production sandbox.

Which fundamentals API is the most accurate?

Accuracy is hard to verify unless the data is traceable, which is why StockFit attaches the SEC accession number and filing date to every value and preserves the amendment trail, so any figure can be checked against the original 10-K, 10-Q, or 20-F on EDGAR. Normalized aggregators (FMP, EODHD, Alpha Vantage, Finnhub) map filings onto their own schema, which is convenient but removes the ability to reconcile a number back to its source. SEC-native providers (Intrinio, Financial Datasets AI, sec-api.io) are closer to the filing but cite at the filing level rather than per value. For work where a wrong number is expensive, prefer an API whose values you can trace.

What is an SEC accession number and why does it matter for financial data?

An accession number is the unique identifier the SEC assigns to every filing on EDGAR (for example, Apple's FY2025 10-K is 0000320193-25-000079). It matters because it is the link between a data point and its source: if a fundamentals API returns revenue of 416.2 billion dollars and attaches that accession number and filing date, you can open the exact 10-K and confirm it. Without the accession trail you are trusting a vendor's normalized number with no way to audit it, and no way to know whether a later restatement silently overwrote the original figure.

What is the difference between as-reported and normalized fundamentals?

As-reported fundamentals preserve the original XBRL concept names and values exactly as the company filed them, including company-specific extension tags, which is ideal for auditing and forensic work. Normalized, or standardized, fundamentals map those issuer-specific concepts onto a stable cross-company schema (revenue, netIncome, epsDiluted, and so on) so a query works identically across thousands of companies. StockFit, FMP, Finnhub, Intrinio, Financial Datasets AI, and sec-api.io all offer an as-reported view; StockFit returns both the as-reported XBRL and the normalized statement, each carrying the same point-in-time filing dates.

Which fundamentals API is best for AI agents?

For AI agents two things matter: a clean, machine-readable schema and citations the agent can surface, since an agent that cannot show its source cannot be trusted. StockFit exposes an MCP server and returns the SEC accession number with every value, so an agent can cite the filing it used. Financial Datasets AI is built for agents and returns a per-filing accession number and filing URL, and FMP, EODHD, Alpha Vantage, and Intrinio also offer MCP servers. The differentiator is not having MCP, which is now common, it is whether each answer carries a source you can verify.

Which fundamentals API is best for backtesting?

Backtesting needs point-in-time data: every value tagged with the date it actually became public, and the original figure preserved when a later filing restates it, otherwise the backtest leaks look-ahead bias. StockFit attaches the filing date and accession number to every value and keeps the amendment trail, so you can reconstruct what was knowable on any past date. Most normalized aggregators return only the latest restated value with no filing date, which is the classic source of look-ahead bias in a backtest.

Do fundamentals APIs cover quarterly and historical financial statements?

Most do, but the depth and the price vary. StockFit, Alpha Vantage, Finnhub, Intrinio, and Financial Datasets AI return annual and quarterly income statements, balance sheets, and cash flow statements with multi-year history; FMP gates quarterly statements and deep history behind higher tiers (its cheapest fundamentals plan is annual-period only); and EODHD covers 30-plus years for major US companies. Always check whether quarterly data and the historical depth you need are included in the tier you are pricing, since several vendors gate history by plan.