Cross-Platform Trends

See which topics are trending at once on Google, Wikipedia, Bluesky and YouTube, with real numbers from each: searches, page views, posts and video views. Where a big Polymarket prediction market exists for a topic, you also get the betting odds and dollars bet.

A topic trending on one site might be noise. A topic people are searching, reading and posting about at the same time is a story. This Actor checks every trend against the other platforms and ranks topics by how many platforms they're trending on.

All sources are official, free and public, so there are no proxies, no logins and no blocking.

Two ways to use it

	Confirmed trends (default)	Full scan
Settings	Your countries, Minimum platforms 2	Countries ALL, Minimum platforms 1
You get	~5–20 high-confidence topics (depends on the news day), each trending on 2+ platforms	~850 trends from ~60 countries and 4 platforms, every one scored with how many platforms confirm it
Run time	~40 seconds	~8 minutes
Good for	Alerts, newsletters, AI agents, "what matters right now"	Dashboards, research, spreadsheets, your own filtering

Every row has platformCount and heatScore, so even the full scan sorts with the strongest trends first.

What it checks

Platform	What you get	Freshness
Google	Trending searches per country, approximate search volume, news stories	Live
Wikipedia	The topic's article, views in the latest hour and day, and how that compares with the week before (e.g. 38x normal)	Hourly views ~2 hours behind; daily views ~1 day behind
Bluesky	Matching trending topic, post count, category	Live
YouTube (optional)	Matching videos from the country's "most popular" chart, with view counts	Live
Polymarket (context)	The matching prediction market among those with the most money bet in 24 hours: dollars bet, the favorite and its odds (e.g. "Gadi Eizenkot 63%")	Live

Polymarket is context, not a trend signal: betting volume shows what bettors care about (elections, football, esports), not what the public is paying attention to. So it never counts toward a topic's platforms or ranking, and never adds topics of its own. It only adds odds to topics already trending elsewhere.

Rows about the same story (sharing a Wikipedia article, Bluesky trend or YouTube video) are merged into one, with the other names listed in relatedTopics. For example, "2026 Ballon d'Or", "Ballon d'Or" and "Harry Kane" become one row.

Topics trending on Bluesky or spiking on Wikipedia that aren't in your countries' Google lists are added as their own rows (country GLOBAL). They're also checked against the Google lists of all major English-speaking countries (US, UK, Canada, Australia, Ireland, India, New Zealand, South Africa), so a story trending on Google in Ireland but not in the US still counts.

By default you only get topics trending on 2 or more platforms: confirmed trends, not noise. Set Minimum platforms to 1 to get everything (see the full scan above).

Sample run (US + UK, October 2026)

#	Topic	Trending on	Evidence	Betting context
1	Noah Wyle named Sexiest Man Alive	Google (IE), Wikipedia, Bluesky	38x normal hourly views; 473 posts
2	khalil mack	Google (US), Wikipedia	1,000+ searches; 1.7x daily views
3	Israel election sees record turnout	Google (NZ), Bluesky	100+ searches; 627 posts	$7.0M bet, Gadi Eizenkot 63% to be PM
4	Nicki Minaj	Wikipedia, Bluesky	144K views (20x normal); 818 posts
5	2026 Ballon d'Or (+ Ballon d'Or, Harry Kane)	Wikipedia, YouTube	452K views (43x); "Ballon d'Or 2026 LIVE", 7.8M views	$19.5M bet, Harry Kane (decided)

Each row also has the top news story, all matched articles and links, and trend tracking:

Field	Meaning
platforms, platformCount	Where the topic is trending right now
heatScore	0–100: how big the topic is compared with the others in the run, across all its numbers. Topics are ranked by platform count, then heat.
isNew, firstSeenAt, hoursTrending, timesSeen	Tracked between runs
newPlatforms	Platforms the topic spread to since the last run, e.g. ["wikipedia"]
relatedTopics	Other names for the same story, merged into this row
googleCountry	Which country's Google list the topic is trending in
polymarketVolume24h, polymarketLeader, polymarket	Betting context: dollars bet in the last 24 hours, the favorite with its odds, and the market link
googleTraffic, wikipediaViews, wikipediaViewsLastHour, wikipediaSpike, wikipediaHourSpike, blueskyPosts, youtubeTopVideoViews	The numbers, flat for spreadsheets
google, wikipedia, bluesky, youtube	Full details from each platform, with links

Download results as CSV, Excel or JSON, or fetch them through the Apify API.

Input

Field	Default	Description
Countries	US	Google trends per country. Each is checked against its own Wikipedia language edition and YouTube chart.
Sources	all	Any mix of Google, Wikipedia, Bluesky and YouTube, e.g. Bluesky + Wikipedia only, plus Polymarket betting context on or off. All free.
Must be trending on	none	e.g. Google: only Google trends that are confirmed on another platform
YouTube API key	empty	Your own free key from Google Cloud Console. Without it YouTube is skipped. Each run uses 1 unit per country of the 10,000 free daily units.
Keywords	none	Only topics mentioning one of these, e.g. a brand, team or person. Headlines, article and video titles are searched too.
Add trends Google doesn't show	on	Bluesky and Wikipedia trends that aren't in Google's list
Minimum platforms	2	Confirmed cross-platform trends only. Set to 1 for every trend from every platform (about 7x more rows, each charged).
Only new topics	off	For alerts
History name, Forget topics after		Trend tracking between runs, as in Google Trending Now

Recipes

• Confirmed-story alerts: schedule hourly with Only new topics on, then connect Slack or email in Apify Console.

Cost: $0.005 per topic saved. A default run saves about 10 topics (~$0.05). For scheduled runs, turn on Only new topics: later runs then save only topics that weren't seen before, so an hourly schedule usually costs cents a day instead of paying again for the same topics every hour.

• Brand or person monitoring: set Keywords (e.g. Nike, Taylor Swift) and Minimum platforms 1 to catch every mention on any platform.
• Social-only trends: set Sources to Bluesky + Wikipedia to see what people are reading and posting about, without Google.
• Spread tracking: schedule hourly and watch newPlatforms to see a topic jump from Bluesky to Wikipedia to Google.
• Content planning: a daily run sorted by platformCount gives the topics people are actively searching, reading and posting about.

Use it with AI agents

Like every Apify Actor, this one works as a tool for AI assistants through the Apify MCP server. Ask your agent "what's trending across platforms right now?" and it can run this Actor and summarize the results.

Only need Google? Google Trending Now is the simpler, cheaper Google-only version.

How "trending" is decided

• Google: in the country's "Trending now" list.
• Wikipedia: the matched article had at least 3x its normal hourly views (and 100+ views) in the latest published hour, or 1.5x its normal daily views (and 1,000+) on the latest day. "Normal" is its average over the 7 days before.
• Bluesky: a current Bluesky trend whose headline mentions the topic.
• YouTube: a video in the country's top 50 whose title, channel or tags mention the topic.
• Polymarket (context only): odds are attached when one of the top 50 markets by 24-hour volume (at least $100,000 bet) mentions the topic in its title or leading outcomes.

Limitations

• Matching uses words, not meaning. "gta 6" matches "Grand Theft Auto VI" (acronyms are handled), but vague trends like "celebrity" or "sunset" may match the wrong article or nothing. Check the wikipedia.article field.
• Wikipedia lags. Hourly views arrive about 2 hours late, so a story that broke an hour ago won't show on Wikipedia yet. Hourly spikes compare with an average hour, so night-time hours read a little low.
• Google's list is short: about 10–25 trends per country.
• Bluesky trends are global and lean English-language and US/UK news.
• Polymarket is a crypto-based prediction market that isn't available to bettors in every country. This Actor only reads its public market data and never places bets. Market odds are crowd estimates, not facts.

Run locally

$python -m src

Input: storage/key_value_stores/default/INPUT.json. Results: storage/datasets/default/.