Free Electricity Data API

Each sample fetches /api/v1/countries/germany.json, prints the latest values of total_generation_twh and demand_twh and the attribution text. Field names and the endpoint come from the OpenAPI description, so they match the files exactly.

curl -s https://worldpowermonitor.com/api/v1/countries/germany.json

curl -s https://worldpowermonitor.com/api/v1/countries/germany.json | jq '.annual | {years: .years[-5:], total_generation_twh: .total_generation_twh[-5:], demand_twh: .demand_twh[-5:]}'
curl -s https://worldpowermonitor.com/api/v1/countries/germany.json | jq -r '.attribution[0].attribution'

import requests

url = "https://worldpowermonitor.com/api/v1/countries/germany.json"
data = requests.get(url, timeout=30).json()

rows = list(zip(data["annual"]["years"], data["annual"]["total_generation_twh"]))
for year, value in rows[-5:]:
    print(year, value)  # total_generation_twh

print(data["attribution"][0]["attribution"])

const res = await fetch("https://worldpowermonitor.com/api/v1/countries/germany.json");
if (!res.ok) throw new Error(`HTTP ${res.status}`);
const data = await res.json();
const { years, total_generation_twh, demand_twh } = data.annual;
console.log(years.at(-1), total_generation_twh.at(-1)); // latest total_generation_twh
console.log(data.attribution[0].attribution);

library(jsonlite)

d <- fromJSON("https://worldpowermonitor.com/api/v1/countries/germany.json")
df <- data.frame(years = d$annual$years, total_generation_twh = d$annual$total_generation_twh, demand_twh = d$annual$demand_twh)
tail(df)
d$attribution$attribution

import pandas as pd, requests

df = pd.DataFrame({k: v for k, v in requests.get("https://worldpowermonitor.com/api/v1/countries/germany.json", timeout=30).json()["annual"].items() if k in ("years", "total_generation_twh", "demand_twh")}).set_index("years")

Country ids (215) are listed in /api/v1/countries.json , US grid region ids (14) in /api/v1/regions.json. An unknown id answers with the site 404 page, not JSON.

Numbers are raw values, never formatted strings; the unit is in the field name (_twh, _mwh, _gw, _pct, _mtco2, _gco2_kwh). A null is a value the source did not report. Every series in a block has one value per entry of its time axis (years, months or dates), in the same order.

Every file carries an attribution field. Keep it, or the line below for each source you use, with any copy or derived work (a chart, an app screen, a dataset). It is the same text as on every dataset page. Ember statistics (countries) and operator data (EIA, ENTSO-E) are separate layers: credit each one where you show it, and do not add figures from the two layers into one number.

Only datasets whose licence allows republishing are served; other series on the site are not available here.

https://worldpowermonitor.com/api/v1/openapi.json is an OpenAPI 3.1 document (43 KB) generated at build time from the code that writes the files: paths, id patterns, response schemas with units, licence and attribution per dataset, cache headers and a real example for every endpoint. It is open to any website (CORS) and has no tracking or external references. This page embeds no third party viewer; import the URL into a tool you trust.

• Postman. Import, choose Link, paste https://worldpowermonitor.com/api/v1/openapi.json, then Continue and Import. It creates a collection with one request per path.
• Insomnia. Create, Import, From URL, and paste the same address.
• Swagger Editor, Swagger UI or Redoc. Open your own copy or the tool's hosted page, and load the same address with its import or URL field.
• Code generation. Point an OpenAPI 3.1 generator at the address to get typed clients. Support for 3.1 varies by tool; if an older tool rejects the document, the endpoint and field tables on this page cover the same ground.

The same address is advertised in a Link: rel="service-desc" response header on this page, as RFC 8631 describes.

• Version. The version is in the path: /api/v1/. These are the v1 files.
• Not breaking. Adding fields, endpoints, datasets, ids or fuel keys; new values with each rebuild (Ember revises past figures, and files always follow the latest release). Ignore fields you do not know.
• Breaking. Removing or renaming a path or field, changing a unit or a type, or changing what a field means. That would ship as /api/v2/. A dataset can also be withdrawn if its licence stops allowing republishing, and its files would then be gone.
• No guarantee. There is no service level agreement, uptime or latency promise. The files are static and served from the Cloudflare edge.
• Caching. Responses send Access-Control-Allow-Origin: * and Cache-Control: public, max-age=300, s-maxage=3600. Files change at most once per site build, so poll no more than hourly.
• Limits. This site sets no quota and issues no keys. Please cache what you fetch.
• Missing values. A null is a value the source did not report. Days of exactly zero regional demand are reported as null, not zero.
• As of dates. Every file states its own as_of and, in changes, what was done to the source data.