Datavidence

Data APIs, built on primary sources

Datavidence

Data you can prove — every value traced back to the record it came from.

We build developer data APIs from authoritative primary sources: normalized into clean schemas, queryable at any point in time, and carrying a citation on every figure — for engineers, analysts, and the AI agents working alongside them.

See it live How we build

The standard

Most data APIs ask you to trust the number. Ours hand you the receipt.

Four commitments hold across every dataset Datavidence ships — whatever the domain.

01 · Source

From the primary record

Built on authoritative filings and registries — not scraped aggregations of aggregations. The data starts where the truth does.

02 · Time

Point-in-time, not just latest

Ask what was known on any date. As-originally-reported figures, so backtests and models stay free of look-ahead bias and silent restatements.

03 · Provenance

A citation on every value

Each figure links to the exact source document and field it was drawn from. Audit it, reconcile it, defend it in review.

04 · Access

AI-native by default

Every API ships as callable tools over MCP as well as REST — so an agent can reach for the data directly, not just your SDK.

The family

One home for data APIs that hold up to scrutiny.

The first is live today. Each new domain inherits the same commitments.

Live

Datavidence Financials

Normalized US-GAAP financial statements from SEC EDGAR — income statement, balance sheet, and cash flow for any US public company, point-in-time, with computed ratios and a source citation on every value. Over REST and MCP.

Explore Datavidence Financials

One value, with its receipt

"revenue": 383,285,000,000,
"as_of": "2023-09-30",
"source": "AAPL 10-K FY2023",
"tag": us-gaap:RevenueFromContract…

Traceable to the filing, not just asserted.

More data domains are in the works — each built to the same standard above.