Skip to main content
One catalog covers the sources, feeds and registries behind the API. GDELT Cloud codes Events, clusters Stories, resolves entity identities and computes analytics from public inputs. Reference datasets retain their originating authority, licence, publication cadence and limitations. Start with /api/v2/search to resolve identity candidates, select the right identity, then send its entity_id as entity= to the relevant endpoint. /api/v2/entities discovers entities in reporting within a date and geography scope. Politicians are people selected by published office evidence (holds_office=true); facilities are candidate type facility with a separate facility_id. New candidate-search examples use country_match=strict for known source country association. A source record can remain unlinked to a resolved entity. Missing country, identity or coverage is unknown, and plan-withheld data is unavailable to that request; neither means zero. An owner relationship and proximity to a facility are different evidence. The Data & coverage page provides dated inventory measurements alongside this dataset reference.
Coverage figures were last measured against production on 2026-07-20. Every limitation below is measured, not estimated — if one stops matching production the number gets fixed, not the wording.

​
What we produce

The core datasets are GDELT Cloud’s own output — we code the events, cluster the stories, resolve the entities and score the tone. The raw article stream they are coded from is an input, not a dataset we resell.
Coded from: GDELT article stream — a raw feed we code from, not a dataset we resell. None of the upstream provider’s own derived products is an input.Every event is coded by GDELT Cloud under our own CAMEO+ taxonomy, not taken from an upstream event feed. History before March 2026 is being backfilled and is incomplete. Event geography is city-level at best — most events resolve to a country or regional centroid, so they should not be treated as point locations.
Coded from: GDELT article stream — a raw feed we code from, not a dataset we resell. None of the upstream provider’s own derived products is an input.Independently coded using ACLED’s published methodology and taxonomy. Not sourced from, endorsed by, or affiliated with ACLED, and not comparable row-for-row with ACLED’s own dataset. Beta.
Coded from: GDELT article stream — a raw feed we code from, not a dataset we resell. None of the upstream provider’s own derived products is an input.Article clusters we build ourselves. They are re-evaluated as coverage arrives, so a story’s article set can change within its first day. The stream selects concrete reportable incidents and material corporate events, including material results, guidance changes and named C-suite succession; it is not a comprehensive corporate-news or routine market-wrap feed.
Coded from: GDELT article stream + public registries — a raw feed we code from, not a dataset we resell. None of the upstream provider’s own derived products is an input.Resolve candidate identities with /api/v2/search, then select an identifier before retrieving its records. /api/v2/entities discovers entities in reporting within a date and geography scope. People, organisations and places are entity types; politicians are people filtered by published office evidence. Facility candidates are available through type=facility and retain their facility_id, separate from an owner entity_id. Missing country association or coverage is unknown, not zero.
Coded from: GDELT article stream + Bluesky posts — a raw feed we code from, not a dataset we resell. None of the upstream provider’s own derived products is an input.Scored by us. Tone is measured only where an entity is actually covered — an absent bucket means no measured coverage, never a neutral reading. Evidence is news coverage plus, for some entity-days, recent Bluesky posts. The model reads both in one pass and returns three readings: the blended headline score, plus news_tone_score and social_tone_score over each source alone. Social is attached only to entities already present in the news stream and never scored alone, but where it is present it is often the majority of the evidence — so read social_tone_score against scored_social_story_count, which is usually a small fraction of the scored total. The split is served per story, per day and per language, and it is FORWARD-ONLY: scores written before 2026-08-22 report null for both components rather than a fabricated zero, and cannot be backfilled. Early access.
Coded from: Our own coded events and stories — our own output, not an ingested feed.Experimental. The API and MCP tools are gated on the Intelligence plan (can_use_intelligence) — they are no longer Admin-only — but the baseline is provisional because the underlying event history only begins in March 2026. Country-level coverage is the binding limit: measured over 107 days, only 2 countries clear the per-country coverage floor on 90+ days and 12 clear it on most days, so every other country is returned as null rather than a fabricated score. Treat world and continent as usable and country as preview. Not derived from, or comparable to, the Federal Reserve’s published GPR index.

​
What we fuse

Multiple public registries joined into one keyed surface and resolved against the entity spine.
About 99% carry coordinates; the remainder are pipelines and oil & gas fields, which are lines and areas rather than points. Roughly half carry an owner resolved to the entity spine — strongest for heavy industry and data centers, weakest for power. Ports carry no owner at all: the World Port Index is a navigational dataset with no ownership layer, so this is a boundary of the source, not a backlog.

​
Reference datasets

Public datasets we ingest, clean, and — where the source supports an identifier — resolve onto the same entity spine as everything else.
Prime awards only. Sub-award chains are not ingested, so exposure through subcontractors is not represented.
Reflects what registrants filed. Non-registration is not observable, so absence is not evidence of no relationship.
US registrants only — a private or foreign-only company has no EDGAR footprint. Supplier and subsidiary relations extracted from filing text are our interpretation of the document, not statements by the filer.
Epoch’s data-center registry is a curated set of the largest known clusters, not a census. Coordinates are geocoded from published addresses, not supplied by the source.
Capacity units differ by tracker — power is MW while heavy industry is tonnes per annum — so capacity is not summable across asset classes.
Current assignments only; fund/unit-of-account codes excluded. No exchange rates or historical legal-tender claims. Unmapped source labels are retained for review and are never guessed.
Annual and lagged; catalog presence does not imply an observed value — a missing observation is served as null, never zero; units differ per indicator and are not summable.
Local currency per US dollar. ECB euro-based quotes are divided by the USD-per-euro quote; this transformation is performed by GDELT Cloud. Original data is freely available from ecb.europa.eu and data.bis.org; BIS data adds no separate charge. Missing currencies stay unavailable. Observation dates are distinct from retrieval dates; neither feed supplies a reliable publication vintage in these responses.
Change against the same month a year earlier, separate from World Bank annual inflation. Original data is freely available at data.bis.org and adds no separate charge. Missing observations remain absent and revisions are retained as retrieval vintages.
A quarterly total, not annual GDP. Source values in millions of dollars are multiplied by 1,000,000. Annual World Development Indicators remain a separate series; publication and observation dates travel with the values.
General government gross debt as a percentage of GDP. It is separate from World Bank central-government debt, whose institutional coverage differs. EU and euro-area aggregates are excluded from country rollups. IMF WEO debt is excluded from commercial publication pending redistribution permission.
Annual and lagged — V-Dem v16 carries a 2025-03-10 vintage and the World Bank indicators trail by one to two years, so these describe a country’s standing condition and can never move on the news. V-Dem indices are expert-coded survey aggregates with published uncertainty intervals; we store the point estimate only. V-Dem publishes no formal licence — its FAQ grants free use for anyone without permission and asks to be cited, which is how these rows are marked.
Series whose upstream terms restrict redistribution are filtered out of every query rather than served with a caveat. GDELT Cloud is not endorsed or certified by the Federal Reserve Bank of St. Louis. Observations cannot predate 1970.
Analytical coverage, not an audit-grade compliance control — use a compliance-grade provider for screening decisions. Lists now include natural persons, served under a minimal person contract: name, aliases, citizenship, Wikidata QID, gender and birth year — never a full date of birth, an identity-document number or an address, which the sources may publish and we withhold. Filter the catalog by category to tell a sanctions list from a debarment or wanted-persons list; only sanctions and export-control memberships count toward the sanctions exposure lens. Country filtering is derived from address text on the US lists and is not complete; two State Department lists carry no country or address at all and cannot be filtered by geography. A list we do not carry returns an explicit error, never an empty result.
Preferred statements without an ended tenure are publisher-asserted current. Normal-rank or ambiguous tenure remains unknown. Published date precision and source links are retained. Complements existing office rosters; not a completeness claim or a compliance-screening service.
Role review distinguishes public officials from support staff; this is not a register of all US officials. The printed report as-of date is an observation, not an appointment date, publication timestamp or proof of continuous current tenure. Reviewed Wikidata identities and exact Wikipedia sitelinks connect roster records to news entities. Compensation, ethnicity and income are not served.
Actor context for geopolitical analysis — NOT a PEP screen and NOT a sanctions-screening tool; never a compliance determination. Dates are VALID time as the publisher states them: an office-holder with no published start date cannot be placed on a date and is excluded from as_of rather than guessed, and status: unknown is served as the source left it, never coerced from a missing end date. Holders reach the news layer only through a Wikipedia identity, so office= on Events and Stories counts a floor, not the roster. The daily inventory is not a completeness claim: holder records are not distinct people, and country presence does not establish a complete national roster. The verified dated load report records source and country inventories, date coverage and resolved-identity coverage; it does not certify complete national rosters or historical coverage.
Terrestrial AIS only — no satellite feed — so coverage is strongest near shore and thins in open ocean. Accrues forward from June 2026; there is no history before that date.
A periodic research release, not a live feed. It ends in 2021 and will not reflect anything more recent until AidData publishes again.

​
Reaching a feed

Each Open Feed is gated by its own entitlement flag. An endpoint your workspace does not carry returns an explicit error naming the flag it needs — never an empty result, which would be indistinguishable from “there is no data for you”. The flag is stated on each family in the API reference, and the error itself is in the error reference.

​
Files, not pages

Coded events are also available as monthly files — Parquet and gzipped CSV, identical rows and column names, each published with a sha256, a byte size and a row count. That is the right shape for backtesting and for loading into your own warehouse, where paging a REST feed is not. Every column is documented in the bulk export schema, and the files themselves are on the Intelligence, Enterprise and Academic plans at Bulk Data Download.