Three ways in
REST API
One endpoint shape for every dataset. JSON over HTTPS, cursor pagination,
versionless URLs.
MCP server
Connect an AI assistant directly to the data. Ask in plain language, get real
figures with sources.
Bulk sync
Whole entities or the whole dataset, for loading into your own store.
Understanding the data
Data model
One object shape for everything. Turkish inflation and Apple’s revenue arrive in
the same envelope.
Data availability
What exists, per country and per dataset. Coverage is uneven and stated plainly.
Trust metadata
Whether a row is fresh, official, and safe to model on.
Units
Rates, indices, balances and levels. The field you cannot skip.
Datasets
Macro indicators
Inflation, growth, labour, rates, consumption, trade. 268 countries.
Company fundamentals
Statements, multiples, filings and insider activity for 1,900 US companies.
FX
ECB reference rates plus returns, realised volatility and trend.
Positioning
CFTC futures positioning and percentile rankings.
Markets
Commodities, credit spreads and rates volatility.
Derived analytics
Real rates, curve spreads and surprise indices we compute ourselves.
First request
Every path is versionless. There is no
/v1 in any documented URL, and none is
required.