Coverage and discovery
The Tauric API serves these data families, each with its own page. Every instrument has one instrument key, which you find by searching.
Routes
| task | route |
|---|---|
| list every served dataset, with its coverage, depth, units, and limitations | GET /v1/datasets |
| read one dataset's record | GET /v1/datasets/{dataset_id} |
| search or list instruments of any kind | GET /v1/instruments |
| one instrument's identity, classification, and datasets | GET /v1/instruments/{key} |
| search or list series | GET /v1/series |
| the classification vocabulary | GET /v1/taxonomy |
Coverage by family
Coverage is per capability: a listing can have quotes and bars and no financials. GET /v1/datasets and
GET /v1/instruments/{key} give the current coverage.
| family | what is served | routes | discover |
|---|---|---|---|
| batch quotes and market views | batch quotes; equity indices, sectors, and movers; rates and credit; fixed-income ETFs; the macro snapshot; cross-asset; commodities; FX; crypto | /v1/quotes, /v1/markets/… | /v1/datasets |
| stocks and ETFs | US-listed stocks, ETFs, and depositary receipts: quote, daily and intraday bars, technicals; profile, analyst consensus, earnings and dividend calendar, corporate actions | /v1/securities/{symbol}/… | /v1/instruments?kind=security&dataset=securities.quote |
| filings and financials | the SEC filings index and financial-statement facts as filed, point in time; Form 4 insider transactions and Form 13F institutional holdings | /v1/securities/{symbol}/filings, /financials, /insider-transactions, /institutional-holders | /v1/instruments?kind=security&dataset=securities.financials |
| economic and market series | economic statistics, index levels, volatility, physical supply and demand | /v1/series/{key}/… | /v1/series (filters, q) |
| rates, credit, and volatility | the yield curve and rates, credit spreads, the macro snapshot, the VIX term structure | /v1/markets/rates/…, /v1/markets/credit/spreads, /v1/markets/macro/snapshot, /v1/series/… | /v1/datasets |
| futures | contract families, their continuous front-month series (quote, bars, technicals), positioning, physical supply and demand | /v1/futures/…, /v1/markets/commodity/… | /v1/instruments?kind=future |
| FX | the 28 maintained G10 pairs: spot quote, daily open/high/low, daily reference rates | /v1/fx/{pair}/…, /v1/markets/fx/pairs | /v1/instruments?kind=fx |
| crypto | exchange spot pairs (quote, daily bars), crypto asset profiles, market overview | /v1/crypto/pairs/{pair}/…, /v1/crypto/assets/{asset}, /v1/markets/crypto/… | /v1/instruments?kind=crypto |
| news and events | articles with their annotations, and semantic search; events extracted from articles | /v1/news…, /v1/securities/{symbol}/events | /v1/news?instrument=… |
| digests, theses, and themes | machine-generated analysis citing its evidence | /v1/digests/market, /v1/digests/{ref}, /v1/securities/{symbol}/digest, /thesis, /v1/theses, /v1/themes | /v1/datasets |
The AI services (research, the copilot, model inference) have their own pages.
Outside coverage, or where a value is not held, the answer is an error code for a whole answer, or a null field (missing values).
The maintained universe
Securities are a maintained set of US listings. GET /v1/instruments?kind=security&dataset=securities.quote lists
them. A registered listing outside the universe gets 404 not_maintained; a ticker that
names no registered listing gets 404 not_registered.
Find an instrument
Every instrument has one key, such as security:AAPL or fx:EURUSD
(instrument keys). Search, then use the key the search
returns:
curl "https://api.tauric.ai/v1/series?q=10-year%20treasury" \
-H "Authorization: Bearer $TAURIC_API_KEY"
curl "https://api.tauric.ai/v1/instruments?q=EUR/USD" \
-H "Authorization: Bearer $TAURIC_API_KEY"
curl "https://api.tauric.ai/v1/instruments?q=future:XCME:ES" \
-H "Authorization: Bearer $TAURIC_API_KEY"
curl "https://api.tauric.ai/v1/instruments/security:AAPL" \
-H "Authorization: Bearer $TAURIC_API_KEY"
/v1/instruments searches every kind; /v1/series searches only series, with the same matching and ordering.
Python examples shows the search from Python.
Query matching
q matches:
- a key: the full key with its kind (
series:…,future:XCME:ES) or its segment, but not the kind word alone; - a ticker;
- a name;
- an everyday name (
PCE_PRICE_INDEXfinds the PCE price index); - an external statistical or exchange code for the instrument (a leading
^is ignored).
A one-word query matches the start of a key token or of a word in a name (cpi, vix), never the middle of one
(es does not find jobless). A query of several words returns only rows where every word matches:
corn ending stocks finds "U.S. Corn: Ending Stocks".
Result order
An exact match of a key, a ticker, or a familiar code ranks first. Ties go by kind: security, future, FX, crypto,
series (BTC lists the bitcoin futures before crypto:BTC-USDT).
Current and historical listings
By default (coverage=current) GET /v1/instruments lists and searches the maintained listings. A listing that
left the universe (departed) or was delisted stays readable, but the default list and q leave it out.
coverage=all includes every registered listing; each row then states maintenance.state (maintained,
departed, or not_maintained) and listing.state (listed or delisted)
(listing lifecycle). coverage changes only securities.
Look up one instrument
GET /v1/instruments/{key} answers an instrument's kind, name, identity, classification, and datasets,
departed and delisted listings included (identifiers). An
unregistered key answers 404 not_registered.
Series lists and the taxonomy
GET /v1/series lists every served series with its series block (series values).
Filter it with kind, asset_class, asset_subclass, and category.
GET /v1/taxonomy lists the classification vocabulary (asset_class, instrument_type, object_kind) and the
content-tag vocabularies (exposure, content_domain) that news filters on
(asset class and instrument type).
The dataset catalog
GET /v1/datasets/{dataset_id} reads one dataset's record. GET /v1/datasets lists the catalog in pages of
complete records: limit (default 10, at most 20) and after page it
(lists). For one instrument, read the datasets of
GET /v1/instruments/{key}. Every field is described in the API reference.
Read these before you rely on a dataset:
values.withheld_fields: the fields null on every answer, with their reason (fx.historynamesclose, so FX bars carry no close).point_in_time.operations: the paths that takeknown_at; when it is empty,known_atis422.limitations: what the dataset does not say, each with the reason codes a response carries for it.