News video reaches 77% weekly: cut platform-native versions

A news team that already has a finished video can publish it as a vertical cut for social feeds and a 16:9 cut for connected TV apps without re-editing, by running Sume's video trim once per version with an output size. The reason to do it now is in the Reuters Institute's 2026 Digital News Report: 77% of people globally consume online news video each week, and most of it is not on the publisher's own site.

Figures below come from the report's executive summary (read 2026-10-02). Sume's side comes from the trim, caption and frames docs. Sume does not source footage, write scripts or fact-check; it only cuts and captions files you supply.

What does the Reuters Institute report say about video?

The summary says a majority of respondents now watch online news video in all 48 Digital News Report markets, and that online video viewing now surpasses broadcast TV news in 45 of the 48. Only Germany, Denmark and the Netherlands keep higher TV news viewership. Growth is concentrated on third-party platforms, while mainstream news organisations' own video consumption fell by 5 percentage points to 23% globally.

It also reports that 27% of respondents now watch on-demand news through apps such as YouTube on their smart TVs. The page does not give a youth video percentage, so this post does not either.

What does that mean for how you export a story?

If viewers meet news on someone else's platform, each platform gets its own version: a vertical cut for phone feeds, a landscape cut for TV apps. Making them by hand is repetitive, but the work is the same each time: pick a range, set a size, burn captions for sound-off viewing.

Video trim takes one media.sume.com clip from your workspace, a start, and one of end or duration. Its optional output sets width and height (256 to 2160) and fps (24, 25, 30 or 60), and only works with precision: exact, the default. A rate of $0.02 per job applies.

How do I produce two versions of one story?

Send two trim requests with different idempotency keys. Each returns a new MP4; the source is untouched. This pair makes a 60 second vertical cut and a 90 second landscape cut of the same piece:

for v in vertical:1080:1920:60 landscape:1920:1080:90; do
  IFS=: read name w h dur <<< "$v"
  curl -X POST https://api.sume.com/v1/video-trim \
    -H "Authorization: Bearer $SUME_API_KEY" \
    -H "Content-Type: application/json" \
    -H "Idempotency-Key: story-1042-$name" \
    -d "{\"video_url\":\"https://media.sume.com/artifacts/artf_demo/story-1042.mp4\",\"start\":0,\"duration\":$dur,\"output\":{\"width\":$w,\"height\":$h,\"fps\":30}}"
done

How do I check the framing and add captions?

The trim docs say output conforms the size on the way out but do not describe how a different aspect ratio is fitted, so check the result instead of assuming. Pull a still with video frames at one or two seconds (unbilled) and look at where faces and lower-third text land.

For sound-off feeds, run video captions on each cut. A standalone caption job is $0.20 for videos up to 60 seconds under the current fixed estimate. Captions come from speech-to-text unless you pass your own script_text or cues, and a silent clip fails as caption_no_speech, so use cues for natural-sound packages.

What stays with the newsroom?

Start with two versions, one vertical and one landscape. Each extra version is another file to caption, check and publish, and at $0.02 per trim and $0.20 per caption job the tooling cost stays under a dollar per story; the real cost is human review of the first and last frame of every cut.

For an end-to-end vertical conversion, see convert horizontal video to vertical and timeline vertical 1080x1920. Track jobs with Jobs and results.

• Editorial judgement: which 60 seconds carry the story.
• Rights to the footage and music in the source file.
• Disclosure rules for AI-made or AI-altered material, which depend on your market.
• Publishing: Sume does not upload to any platform.