Global News Event Signals — GDELT Records
Pricing
from $33.50 / 1,000 news event records
Global News Event Signals — GDELT Records
Machine-coded global news event signals from GDELT: event type (CAMEO), participant countries, tone, place, mention counts and source URLs. Filter by event type, participant country, minimum mentions and event date. Event records, not articles. No person names. Pay per record.
Pricing
from $33.50 / 1,000 news event records
Rating
0.0
(0)
Developer
NexGen Signal
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
2 months ago
Last modified
Categories
Features
Share
Open source data. No API key. Pay per record.
This actor turns the GDELT Project's global news event stream into clean, per-event signal records. Each run reads a bounded number of GDELT's public 15-minute event files and delivers one record per coded event — the event type (CAMEO), the tone, the mention/source/article counts, the action geography, the source article URL, and the org/state actor codes. Person-name actor fields are excluded at ingest, so no individual is named in a record.
What this is, in one paragraph
GDELT machine-codes the world's news into structured events every 15 minutes. This actor reads those event files directly and flattens each event into a row you can load into a spreadsheet, a database, or a pipeline: what kind of event, how it was toned, how widely it was reported, where it happened, and a link to a source article. It emits the CAMEO actor codes (country / type / known-group) — never the free-text actor names, which can be individuals. Values are verbatim; nothing is scored or re-derived.
Who buys this and for what job
- Risk, geopolitical, and macro teams tracking event signals (protests, statements, conflict, cooperation) by geography and tone.
- Media-intelligence and research teams who want a structured event feed with source links.
- Quant and data teams feeding tone and event counts into models on a bounded, repeatable window.
- Data engineers wanting a keyless, one-call event feed.
Person suppression (absolute) and bounded windows
Two hard rules are built into the code:
- No person names. GDELT's Actor1Name / Actor2Name columns carry free-text names that can be natural persons. This actor never reads those columns — only CAMEO code fields are emitted, so a person's name cannot enter a record. Verified by the acceptance suite (a row with person names in those columns is ingested and the emitted record contains no name).
- Bounded windows only. A run reads at most
maxFilesrecent 15-minute files (hard cap 96 = one day). There is no unbounded backfill — the window is bounded in code.
Sample output

Real rows from a live run of this actor (first 5 rows, selected columns).
One full record from the same run, exactly as delivered:
{"record_id": "1324917981","event_date": "20250925","year": "2025","actor1_code": "BUS","actor1_country": null,"actor1_known_group": null,"actor1_type1": "BUS","actor2_code": "SGPGOV","actor2_country": "SGP","actor2_known_group": null,"actor2_type1": "GOV","event_code": "036","event_root_code": "03","quad_class": "1","goldstein_scale": "4.0","num_mentions": "1","num_sources": "1","num_articles": "1","avg_tone": "2.44897959183673","action_geo_fullname": "Singapore","action_geo_country": "SN","action_geo_lat": "1.366667","action_geo_long": "103.8","date_added": "20260925173000","source_url": "https://www.thehindu.com/news/national/karnataka/singapore-firms-to-expand-presence-make-additional-investment-of2000-cr-in-state-mb-patil/article71509950.ece","source": "GDELT Project (GDELT 2.0 Events)","licence": "The GDELT Project — open data, free to use with attribution to The GDELT Project. Reformatted to event-signal records; values verbatim. Person-name actor fields excluded at ingest.","attribution": "Source: The GDELT Project","disclaimer": "Machine-coded global news event signals from the GDELT Project (automated event extraction with inherent noise). These are event signals, not verified facts and not a person database. Source articles are linked; verify before relying.","observed_at": "2026-09-25T17:23:20Z"}
Pricing
| Event | Free plan | Bronze | Silver | Gold / Platinum / Diamond |
|---|---|---|---|---|
news_event_record | $0.05 | $0.045 | $0.04 | $0.0335 |
Delivered before charged. Blocked runs and empty windows cost nothing.
Input
| Field | Type | Default | Description |
|---|---|---|---|
eventCodes | array | — | Optional. CAMEO event codes or their first digits (14 protest, 19 fight, 190 use of conventional force, 04 consult). |
actorCountries | array | — | Optional. CAMEO 3-letter country codes (USA, CHN); either participant may match. |
minMentions | integer | — | Optional. Keep events with at least this many mentions (num_mentions). |
eventLastDays | integer | — | Optional. Keep events whose event_date is within the last N days (recent files also carry older events). |
maxFiles | integer | 3 | Recent 15-minute files to read (newest first). Hard cap 96 (one day). |
actionCountry | string | — | Optional FIPS country code (US, UK, CH…) to keep only events in that country. |
maxRecords | integer | 10 | Ceiling on records delivered and billed. |
Each 15-minute file holds ~1,400 events, so the default (3 files) is a bounded sample of a few thousand events.
Example: {"eventCodes": ["14"], "minMentions": 5, "eventLastDays": 2, "maxFiles": 12, "maxRecords": 10} returns protest events from the last two days, mentioned at least 5 times, found in the last ~3 hours of GDELT files. These are machine-coded event records (who did what to whom, where), not articles or headlines: there is no keyword search of article text. Two dates: event_date is the day the event happened as GDELT coded it; date_added is when GDELT processed the article (UTC). A recent file can carry an older event (seen on 30 Sep 2026: events dated 2016 and 2025 in that day's files), so use eventLastDays or check event_date before treating a record as news of the day. Country codes differ by field: actorCountries uses CAMEO 3-letter codes for participants; actionCountry uses FIPS 2-letter codes for where it happened.
Output
One JSON object per event. Example:
{"record_id": "1210034567","event_date": "20260818","year": "2026","actor1_code": "USA","actor1_country": "USA","actor1_type1": "GOV","actor2_code": "BUS","event_code": "046","event_root_code": "04","quad_class": "1","goldstein_scale": "7.0","num_mentions": "4","num_sources": "1","num_articles": "4","avg_tone": "0.307","action_geo_fullname": "Washington, District of Columbia, United States","action_geo_country": "US","source_url": "https://example.com/news-article","source": "GDELT Project (GDELT 2.0 Events)","licence": "The GDELT Project — open data ...","attribution": "Source: The GDELT Project","disclaimer": "Machine-coded global news event signals ... not verified facts and not a person database ...","observed_at": "2026-08-18T00:00:00Z"}
An unbilled RUN_RECEIPT records the robots check, the latest slot, files read, events seen, delivered, and whether charged equals delivered.
Cost guidance
One news_event_record per delivered event: cost = events pulled × your plan's per-record price. 1,000 events is $50 (Free) / $33.50 (Gold).
Honest limitations
- Machine-coded signals, not verified facts. GDELT's automated coding has inherent noise. Use the source URL to verify; these are signals, not confirmed events.
- Event records, not the theme/entity graph. This is the GDELT Events feed (event codes, tone, geography, counts). It does not include GDELT's GKG theme/person/organisation graph — deliberately, to keep person names out.
- Actor codes, not names. Actor fields are CAMEO codes (country/type/known-group), never free-text names.
- Bounded window. Recent files only, capped; not a historical backfill.
Differentiation
This is a per-event signal record from GDELT's open event stream — a structured, source-linked event feed. It is not a brand/social monitor or an alerting service, and (unlike the NexGen Signal statistical actors) it is a news-event signal, not an official government statistic.
Input fields
Every input this actor accepts, read from its input schema:
| Field | Type | Default | Description |
|---|---|---|---|
maxFiles | integer | 3 | How many recent GDELT 15-minute event files to read (newest first). Bounded window — max 96 (one day). More files = more events. |
actionCountry | string | Optional FIPS country code (e.g. US, UK, CH) to keep only events whose action geography is that country. | |
maxRecords | integer | 10 | Maximum records delivered and billed. You are billed only for records delivered. |
Default changed on 30 Sep 2026: if you leave maxRecords out, a run now returns up to 10 records (it was 500). Set maxRecords yourself to get more — the maximum is unchanged.
Output fields
Every field in each delivered record, read from its output schema:
| Field | Type | Meaning |
|---|---|---|
record_id | string | GDELT GlobalEventID. |
event_date | string | Event day (YYYYMMDD). |
year | string | Event year. |
actor1_code | string | CAMEO actor-1 code (org/state code, not a person name). |
actor1_country | string | Actor-1 country code. |
actor1_known_group | string | Actor-1 known-group code. |
actor1_type1 | string | Actor-1 type code. |
actor2_code | string | CAMEO actor-2 code. |
actor2_country | string | Actor-2 country code. |
actor2_known_group | string | Actor-2 known-group code. |
actor2_type1 | string | Actor-2 type code. |
event_code | string | CAMEO event code. |
event_root_code | string | CAMEO root event code. |
quad_class | string | Event quad class (1-4). |
goldstein_scale | string | Goldstein scale, verbatim. |
num_mentions | string | Number of mentions. |
num_sources | string | Number of sources. |
num_articles | string | Number of articles. |
avg_tone | string | Average tone, verbatim. |
action_geo_fullname | string | Action geography full name. |
action_geo_country | string | Action geography country code. |
action_geo_lat | string | Action geography latitude. |
action_geo_long | string | Action geography longitude. |
date_added | string | GDELT date added (YYYYMMDDHHMMSS). |
source_url | string | Source article URL. |
source | string | Data source (GDELT). |
licence | string | Licence and reformatting notice. |
attribution | string | Required attribution string. |
disclaimer | string | Machine-coded event-signal disclaimer. |
observed_at | string | UTC retrieval timestamp. |
FAQ
How much does it cost? You pay $0.05 per news event record delivered to your dataset (less on higher Apify plans). Set the record limit in the input to cap what a run can cost. Records that are not delivered are never charged.
How do I get the data out? Every run saves its records to an Apify dataset. You can download it as JSON, CSV, Excel, XML or HTML from the run page, or read it through the Apify API.
Can I run it on a schedule? Yes. Use Apify Schedules to run it daily, weekly or monthly, with the same input each time.
Can I call it from my own code? Yes. Call nexgensignal/global-news-event-signals through the Apify API or the Apify Python and JavaScript clients, pass the same JSON input as the form, and read the dataset when the run finishes.
What happens if the source is down? The run ends with a clear status message, and nothing is charged for records that were not delivered.
The NexGen Signal family
Part of the NexGen Signal family of official-source, pay-per-record products:
- Eurostat House Price Index
- Eurostat Waste Generation
- Eurostat Waste Treatment
- Eurostat Business & Consumer Confidence
- Eurostat Income Distribution (EU-SILC)
- Eurostat Construction Production
- Eurostat Immigration Statistics
- Eurostat Emigration Statistics
- Eurostat Road Freight Transport
- Organization Sanctions & LEI Risk
Source: The GDELT Project (open data, free to use with attribution). Reformatted to event-signal records; values verbatim. Person-name actor fields excluded at ingest. Machine-coded signals — not verified facts, not a person database.