GDELT CloudGDELT Cloud
Snapshot assembled 2026-10-15 04:30 UTC

Data Status

What the GDELT Cloud API can serve right now, and how it got there. The top section reports what /api/v2 will actually return, so every figure there is something you can retrieve. The pipeline section answers a different question: is coding running right now.

Coded events since
2026-03-01
Clustered stories since
2026-03-08
Current through
2026-10-15

Every measurement below is a 30-day, 7-day or 48-hour window. These three dates are the whole envelope. Records dated before them exist — a 2026 article about a historical incident writes a back-dated event — but they are individual records, not coverage of those periods. Every /api/v2 list and summary response carries the same envelope as meta.coverage.

Available via the API

Read from serving_events and serving_stories — one settled row per record, which is exactly what /api/v2/events and /api/v2/stories return. Windows are whole closed UTC days; today is excluded because it settles on the hour and is complete only in arrears.

Events · last 30 closed days
111.1k
settled and retrievable
Stories · last 30 closed days
420.0k
settled and retrievable
Events settled through
2026-10-14
last settle 4h 1m ago
Stories settled through
2026-10-14
last settle 2h 43m ago

Events available per date · last 30 closed days

09-1509-3010-14

Stories available per date · last 30 closed days

09-1509-3010-14

Core API latency · last 30 days, per day

Response time for the Core endpoints — Events, Stories, Entities and Search — measured on successful API-key requests. Rejected and failed requests are excluded. Hover any bar for that period’s exact p50 and p95.

2026-09-152026-10-15
p50 (median)p95

Core API latency · last 48 hours, per hour

2026-10-13T04:00Z2026-10-15T04:00Z
p50 (median)p95

Coding pipeline · live warehouse

The processing clock: what the coder has produced in the last 60 minutes. These numbers run ahead of what the API serves by design — newly coded events become retrievable on the next refresh.

Most recent event
2026-10-14 15:41:19 UTC
coded_at
Events coded · last 1h
0
rolling 60 min
Events coded · last 24h
2.5k
rolling, processing clock
Events coded · last 30d
113.1k
includes re-codes of older dates
Stories formed · last 24h
9.2k
content clock — time_bucket

Events coded per clock hour · last 24h

Buckets are whole UTC clock hours on coded_at. The tile above is instead a rolling 60 minutes, and coding lands in hourly batches — so between batches the rolling window can sit well below the last full hour, and just after one it can sit above.

04:0016:0004:00

Stories formed per hour · last 24h

A content clock: time_bucket is the UTC hour a cluster was formed from its articles. This chart groups Stories by content time; it does not measure current processing delay or when each Story first became available through the API.

04:0016:0004:00

Event metric distributions

Over the settled events above, bucketed by whole closed UTC days. CAMEO+ detail metrics only count CAMEO+ events with scores; Goldstein metrics count Conflict and CAMEO+ Political events. Values are the ones materialized at settle time — the same numbers the API returns.

Last 7 settled days

20.6k events

Significance

Composite 0-1 rank combining intensity, article support, fatalities, and confidence.

20.6k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Confidence

Model confidence for the structured event record.

20.6k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Goldstein scale

Signed Conflict and CAMEO+ Political valence from -10 to 10.

11.7k
-10--8
-8--6
-6--4
-4--2
-2-0
0-2
2-4
4-6
6-8
8-10

Goldstein severity

Absolute Goldstein intensity from 0 to 10, regardless of positive or negative valence.

11.7k
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10

Magnitude

CAMEO+ detail metric from 0 to 10.

19.0k
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10

Systemic importance

CAMEO+ detail metric from 0 to 1.

19.0k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Propagation potential

CAMEO+ detail metric from 0 to 1.

19.0k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Market sensitivity

CAMEO+ detail metric from 0 to 1.

19.0k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Last 30 settled days

111.1k events

Significance

Composite 0-1 rank combining intensity, article support, fatalities, and confidence.

111.1k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Confidence

Model confidence for the structured event record.

111.1k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Goldstein scale

Signed Conflict and CAMEO+ Political valence from -10 to 10.

61.0k
-10--8
-8--6
-6--4
-4--2
-2-0
0-2
2-4
4-6
6-8
8-10

Goldstein severity

Absolute Goldstein intensity from 0 to 10, regardless of positive or negative valence.

61.0k
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10

Magnitude

CAMEO+ detail metric from 0 to 10.

103.4k
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
9-10

Systemic importance

CAMEO+ detail metric from 0 to 1.

103.4k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Propagation potential

CAMEO+ detail metric from 0 to 1.

103.4k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Market sensitivity

CAMEO+ detail metric from 0 to 1.

103.4k
0-0.1
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.6
0.6-0.7
0.7-0.8
0.8-0.9
0.9-1

Taxonomy and coverage · last 7 closed days

Every table below counts settled records, so each figure is reproducible from a /api/v2 query over the same window. The denominators are 20,580 events and 86,824 stories.

Events by family

  • CAMEO+
    18,964
  • Conflict
    1,616

Top 15 countries

Display names as the API serves them, from the settled geo column.

  • United States
    1,916
  • India
    1,436
  • Nigeria
    557
  • China
    477
  • Italy
    472
  • Ukraine
    455
  • United Kingdom
    455
  • Russia
    439
  • Spain
    386
  • Greece
    363
  • Turkey
    363
  • Taiwan
    343
  • Mexico
    340
  • France
    312
  • Brazil
    310

Top 15 CAMEO+ subcategories

  • Make public statement
    2,511
  • Criticize or denounce
    783
  • Fiscal Policy Action
    552
  • Engage in negotiation
    510
  • Appeal
    373
  • Praise or endorse
    365
  • Police investigation or operation
    338
  • Appeal for policy change
    326
  • Express intent to change policy
    250
  • Consider policy option
    247
  • Make a visit
    245
  • Production capacity change
    240
  • Consult
    238
  • Host a visit
    206
  • Transport disruption
    171

Top 10 Conflict subcategories

  • Peaceful protest
    617
  • Air/drone strike
    227
  • Attack
    133
  • Protest with intervention
    107
  • Arrests
    102
  • Armed clash
    89
  • Shelling/artillery/missile attack
    86
  • Disrupted weapons use
    68
  • Other
    45
  • Violent demonstration
    37

Top 15 story cluster categories

  • CAMEO+ · Political
    35,596
  • CAMEO+ · Crime
    13,235
  • Conflict · Security
    11,823
  • CAMEO+ · Economic
    6,982
  • CAMEO+ · Infrastructure
    5,711
  • CAMEO+ · Environment
    4,294
  • CAMEO+ · Corporate
    3,131
  • CAMEO+ · Health
    2,379
  • CAMEO+ · Technology
    2,020
  • CAMEO+ · Information
    999
  • CAMEO+ · Demographic
    654

Top languages · across served stories

Article-languages attached to a served Story: 169,121 over this window. This is NOT the ingest corpus below — it counts only the articles that made it into a Story you can retrieve, 4.2x fewer.

  • English
    54,452
  • Spanish
    19,730
  • Arabic
    10,647
  • Russian
    10,008
  • Greek
    7,692
  • Chinese
    4,750
  • French
    4,636
  • Portuguese
    4,511
  • Italian
    3,771
  • Ukrainian
    3,628
  • Albanian
    3,535
  • Turkish
    3,043
  • Romanian
    3,009
  • German
    2,407
  • Bulgarian
    2,346
  • Serbian
    2,214
  • Lithuanian
    2,201
  • Bosnian
    1,956
  • Czech
    1,764
  • Nepali
    1,490

Top languages · ingested article corpus

Every article we language-tagged: 709,046 over the same window, whether or not it ended up in a served Story. Whole closed UTC days — the old "last 24h" label spanned 24 to 48 hours depending on the time of day.

  • English
    255,341
  • Spanish
    83,699
  • Arabic
    36,645
  • Russian
    36,355
  • Greek
    27,280
  • Chinese
    22,623
  • French
    19,827
  • Italian
    18,894
  • Portuguese
    17,347
  • Turkish
    12,239
  • German
    12,134
  • Albanian
    11,646
  • Ukrainian
    11,032
  • Romanian
    10,426
  • Serbian
    8,128
  • Lithuanian
    8,013
  • Bulgarian
    7,950
  • Bosnian
    7,627
  • Korean
    6,377
  • Czech
    5,948

How this data is produced: GDELT Cloud codes all structured events (CAMEO+ and Conflict) and clusters Stories ourselves, using frontier LLMs against an article stream. The structured events, story clusters, and event/story metrics above are our own product — not redistributed third-party records.

Two clocks, kept apart: a content clock (the date an event or story belongs to) and a processing clock (when we coded it). What the API serves is organised by the content clock, because that is what a query filters on; the pipeline section is on the processing clock. Numbers from the two are not interchangeable and are never mixed inside one panel.

Upstream sources: raw article feed from The GDELT Project; energy infrastructure asset records from Global Energy Monitor.