Multimedia & videoBy country.

The MMED category covers video sharing, streaming, podcasts, and image-heavy social platforms. Blocking these is a common middle-ground tactic: it cuts off the most viral political content without the international backlash of blocking news outlets directly.

This editorial context explains the subject. The evidence-volume index does not establish intent, sequence, current restrictions or national scope.

These are examples from the original topic guide, not findings that these domains are blocked now or everywhere.

MMED groups multimedia domains. The response does not verify every app, video stream or content category on a platform.

The checked-in topic query counts all evidence rows assigned to one Citizen Lab category and groups them by country. It does not filter those rows by a blocking signal or observation date. Evidence volume therefore does not establish a confirmed block, intent or national coverage.

Relative index: round(country evidence count × 100 ÷ largest country count in this category). It is a ratio to the maximum, even though the API names its scoring method “percentile_against_max_in_category”. Scores are not comparable across topics as a measure of censorship severity. A rounded score of zero can still have evidence rows.

Source bands: severe ≥70, high ≥50, medium ≥25, low ≥10, free below 10 in the checked-in handler. These names are retained as source labels. “Free” is not a finding that a country permits access; “severe” is not independent confirmation of broad censorship.

Domains: the response field named blocked contains frequent domains in the category evidence query. It does not prove that each listed domain is blocked. The endpoint currently returns up to five per country; all returned values appear here.

Trend: the checked-in handler always emits “stable”. This preview retains that field in each record without drawing a movement chart or implying that conditions were unchanged.

Time: generated_at_utc and dateModified are set when the endpoint builds its response. They are not the latest measurement date. No observation window is inferred; source caching for five minutes is not a promise of new evidence every five minutes.

The source ecosystem includes OONI, Censored Planet, IODA and Voidly probes. This category response does not identify every contributing source. Domain categories follow the Citizen Lab test lists; example domains and returned domains have different meanings.

Voidly original material: CC BY 4.0. Upstream source terms remain applicable. Full methodology ↗

The topic source returned HTTP 503. No date or zero-value summary has been substituted.

The public JSON endpoint does not require an API key. Its current cache header is five minutes; this is not a guarantee of new measurements.

curl 'https://api.voidly.ai/data/censorship-index.json?topic=multimedia'

const response = await fetch('https://api.voidly.ai/data/censorship-index.json?topic=multimedia')
if (!response.ok) throw new Error('Topic source unavailable')
const data = await response.json()
console.log(data.countries)

The original guide supplied the iframe below. The current widget reads country, theme and size, but ignores topic and defaults to a country view. This snippet is retained for reference; it does not currently deliver a topic ranking. No iframe is loaded here.

<iframe
  src="https://voidly.ai/widget/embed?topic=multimedia"
  width="100%"
  height="500"
  frameborder="0"
  title="Voidly censorship: Multimedia & video">
</iframe>

Use the working category API above for a topic-specific integration. The existing country widget has a measurement option; review its settings before embedding.