Google Trends Keyword Interest Scraper
Pricing
from $3.50 / 1,000 keyword reports
Google Trends interest over time for any keywords: weekly/hourly series, average and peak interest, interest by region and related queries. Uses only Google endpoints verified live, with cookie warm-up and proxy rotation so runs don't silently fail.
Search terms to pull Google Trends interest for. Each keyword becomes one result row with its full interest-over-time series.
[ "apify", "web scraping"]How far back to pull interest data. Longer ranges return coarser buckets: 'today 12-m' is weekly, 'today 5-y' is monthly, 'now 7-d' is hourly - exactly as Google itself aggregates them.
Two-letter country code (US, GB, IN), a sub-region (US-CA, GB-ENG), or leave empty for worldwide.
Off (default): every keyword gets its own 0-100 scale, so unrelated keywords stay readable. On: keywords are scored relative to each other in groups of 5 (Google's limit), which is the only way to tell which term is genuinely bigger - but a small term next to a huge one will read as near-zero.
Which Google property the interest is measured on.
Google Trends category id to narrow the topic, or 0 for all categories. Ids come from Google's own category tree (for example 7 = Finance, 71 = Food & Drink, 13 = Computers & Electronics).
Add the regional breakdown (states/countries ranked by relative interest) to each result. Costs one extra request per keyword group, no extra charge.
Add Google's top and rising related queries to each result. Google publishes no related queries for low-volume keywords, in which case these come back as empty lists.
Cap on the regional breakdown, highest interest first. Regions with zero interest are always dropped.
Cap on each of the top and rising related-query lists.
Safety cap on how many keywords one run processes, so an oversized list cannot run up a bill. Extra keywords beyond this are ignored. Google throttles bursts, so large runs are slow: roughly 25 keywords in 3 minutes, 100 in 20 minutes when each keyword is scraped on its own scale. Raise the run's timeout before raising this.