Dailymotion exposes a trending endpoint that requires no API key. I pulled 100 videos from it on October 7, 2026 — two pages of fifty — and then did the analysis that the endpoint itself does not do: I counted who was actually publishing.
The result is uncomfortable for anyone who treats a "trending" feed as a proxy for platform-wide activity.
What the endpoint returns
Two pages, fifty entries each. Every entry carries an id, a title, an owner, an owner id, a channel, a duration in seconds, a thumbnail, and a view count. No authentication, no rate limiting encountered at this volume.
One thing worth knowing before you build on it: the channel parameter does not work the way the documentation implies. Passing channel=news or channel=sport with a sort modifier returns a 400. The reliable path is to pull the feed whole and filter locally, which is what I did.
The concentration
Twelve distinct accounts produced the hundred entries.
43 videos, one account
18 videos, a second
9 videos, a third
8 videos, a fourth
6 videos, a fifth
5 videos, a sixth
2 videos, a seventh
1 video each, five more accounts
Four accounts account for 78 entries out of 100. Six accounts account for 89.
If you had built a recommendation system on the assumption that this feed represents a broad sample of what the platform serves, the assumption fails here. This is closer to a syndication list than a ranking.
The format those entries share is visible in any two of them. A 54-second micro-drama teaser, 77 views:
https://www.dailymotion.com/video/xbhty7m
A second teaser from the same cluster, same length, 33 views:
https://www.dailymotion.com/video/xbhrgti
And the standout, from a different account entirely: a ten-minute news documentary that drew the highest view count in the hundred at just over 99 thousand plays, roughly 1,300 times the teasers above.
https://www.dailymotion.com/video/x9o94sq
The channel label tells you the same thing
44 of the 100 entries sit in the shortfilms channel. On Dailymotion that label is not a studio category, it is where user-uploaded micro-drama lands. Nearly half the feed carries it.
Which means the feed's apparent diversity, sold as a ranking of videos from across the platform, is largely one format from one region of the catalogue, contributed by a small set of accounts.
Two entries from outside that cluster make the contrast concrete. A seven-minute piece, 704 views:
https://www.dailymotion.com/video/xbfge22
A five-and-a-half-minute motorsport recap, 7 views:
https://www.dailymotion.com/video/xbhp1j2
Both sit in the same middle band the feed barely populates at all.
Freshness, and why it matters
96 of the 100 entries were published within seven days of the pull. The oldest entry in the set was 2.8 days old.
A trending feed that is 96 percent under a week old is not ranking accumulated interest. It is ranking what got traction this week, inside a very small set of publishers, in one format.
What this means if you consume the feed as a signal
Two failure modes, both easy to fall into.
Treating it as platform-wide: it is not. It is a rotating window, and this week the window is largely one account.
Treating the ranking as a merit order: it is not that either. The single most-viewed entry in the hundred was a ten-minute news piece from a different account entirely, which beat every micro-drama entry by a wide margin. The feed listed it, but nothing about the ordering suggested it was the outlier it is.
The useful move is to pull the feed, group by owner, and look at the distribution before you look at the list.
Method
Collected from Dailymotion's public API on October 7, 2026, two pages of fifty, no key. Fields as returned. Account names shortened here for readability; the full grouped breakdown is below.
Full dataset, group by owner, with every video linked: https://vidmotionsave.com/
Top comments (0)