Datasets
Discover dataset implementation and account-specific readiness, supported filters and point-in-time availability. Default order: id ascending. Results are limited to the entitled scope. Never 501 or 503: a dataset with nothing published is reported as rows, implementation_status planned or availability unavailable in /v1/datasets and coverage status unavailable in /v1/datasets/{dataset}/coverage. Access requires the operation capability on the API key.
/v1/datasets- Required capability
- reference.read
- Data timing
- As published by the source
- Credits
- 5 credits per request.
Parameters
as_ofstringoptional · query- Not supported: only the current catalog is served, and any as_of returns 422 point-in-time-unavailable. Omit it.
limitintegeroptional · query- Maximum records per page.Default:
100 cursorstringoptional · query- Opaque cursor, expires after one hour; binds filters, order, resolved as_of, account/key and entitlement. Repeat the original query unchanged.
formatstringoptional · query- Response format. csv returns the data rows with a header row, for spreadsheet functions such as IMPORTDATA; nested fields become dotted columns and lists one JSON array cell, the next page cursor is returned in X-Next-Cursor and meta.errors in X-Meta-Errors. A result without rows is the header row alone, listing the documented fields of a row. Default json, or csv when format is omitted and the Accept header ranks text/csv above JSON; a format the Accept header does not admit returns 406 not-acceptable.
jsoncsv
Response fields
Returns data as an array of records. All example values are illustrative.
idstring- Dataset slug.
namestring- Dataset name.
descriptionstring- Scope.
resource_pathsarray- resource paths
implementation_statusstring- planned: no data source is connected yet, so the dataset's endpoints return 501 dataset-unavailable and its description starts with "Not yet available" (x-availability unavailable in the contract). implemented: a source is connected; availability reports readiness for this account.
plannedimplemented availabilitystring- Delivery readiness for this account.
unavailablepartialready supports_point_in_timeboolean- supports point in time
earliest_as_ofstring · null- First UTC date as_of can name for this dataset and key: the day its knowledge history starts. An earlier as_of returns 422 point-in-time-unavailable naming this date. Null when the dataset keeps no knowledge history or has nothing published.
filtersarray- filters
capabilitystring- Required capability.
Response metadata
Responses include meta.timezone. Dataset responses include next_cursor; health and exports have their own metadata. Request tracing uses the X-Request-ID header.
When something goes wrong
An empty data: [] is a successful query with no matching records. Invalid symbols, permissions and malformed parameters produce an RFC 9457 problem with an actionable detail.
Plans
Requires reference.read. Access, recency and history depth come from the plan definitions the API enforces.
| Plan | Access | Recency | History |
|---|---|---|---|
| Free | Included | As published | Not date-bounded |
| Developer | Included | As published | Not date-bounded |
| Pro | Included | As published | Not date-bounded |
| Commercial | Included | As published | Not date-bounded |
| Scale | Included | As published | Not date-bounded |
Free requests are limited to AAPL, NVDA, JNJ, PEP, MSFT, SPY, QQQ.
History is how far back a plan may request. How far records actually go for a symbol is history_from in dataset coverage. Nothing is real-time. Compare plans