Google Trends Scraper — Compare Keywords, Interest & Trending Now

Get Google Trends data as clean JSON — no browser, no API key, no login. Three things in one Actor:

1. Keyword reports — how interest in a search term changed over time (0–100), which regions search for it most, and which related queries are rising.
2. Keyword comparison — 2–5 terms in ONE query on a shared 0–100 scale, exactly like the compare view on trends.google.com. This is the feature most Trends scrapers get wrong: if you scrape terms separately, Google normalizes each one to its own scale and the numbers are NOT comparable.
3. Trending Now — the searches trending right now in any country, with search volume, growth %, start time, related keywords, and news articles.

Why use this one?

• Fast. Trending Now returns in seconds; a full keyword report typically takes under a minute. No headless browser anywhere.
• Comparison that actually works. compareKeywords is a plain input field — no fragile flags — so it behaves the same from the Console, the API, and AI agents (MCP).
• You only pay for data. A failed keyword, an empty result, an error — all free. If the run charges you, it delivered data.
• Built for Google's rate limits. Google throttles Trends hard, per IP. This Actor rotates a fresh residential IP on every data request and retries automatically, which is why runs succeed where scripts and cheaper actors return nothing.

What this Actor does NOT do

Honesty up front:

• No absolute search volumes for keywords. Google Trends only publishes relative interest (0–100). Only Trending Now items carry approximate real volumes (e.g. 200000) — because Google publishes those.
• No related topics. Google returns empty topic lists to any automated session, so this Actor doesn't pretend to offer them. Related queries work fully.
• Comparison is capped at 5 terms — that's Google's own limit for a shared scale. Individual keywords are capped at 20 per run to keep runs reliable.

Input examples

What's trending in the US right now:

{
    "scrapeTrendingNow": true,
    "trendingGeos": ["US"]
}

Full report for each keyword:

{
    "keywords": ["standing desk", "ergonomic chair"],
    "timeframe": "today 12-m",
    "geo": "US"
}

Compare terms on one scale:

{
    "compareKeywords": ["notion", "obsidian", "evernote"],
    "timeframe": "today 3-m",
    "geo": ""
}

All three can be combined in one run. geo empty = worldwide. Timeframes: now 1-H, now 4-H, now 1-d, now 7-d, today 1-m, today 3-m, today 12-m, today 5-y, all, or a custom range like 2024-01-01 2024-12-31.

Output examples

Trending item (one per trend):

{
    "type": "trending_now",
    "trending_geo": "US",
    "keyword": "taylor farms",
    "search_volume": 200000,
    "volume_growth_pct": 1000,
    "started_at": "2026-07-16T21:20:00.000Z",
    "is_active": true,
    "trend_keywords": ["taylor farms", "taylor farms lettuce"],
    "topics": ["Food and Drink", "Health"],
    "news": [
        {
            "title": "Cyclospora Linked to Taylor Farms Lettuce Sent to Taco Bell",
            "url": "https://www.nytimes.com/...",
            "source": "The New York Times",
            "picture": "https://encrypted-tbn1.gstatic.com/..."
        }
    ]
}

Keyword report (one per keyword):

{
    "type": "keyword",
    "keyword": "coffee",
    "geo": "US",
    "timeframe": "today 12-m",
    "stats": { "average": 75.4, "peak": 100, "peak_date": "2026-04-12", "latest": 76, "direction": "rising" },
    "interest_over_time": [{ "date": "2025-07-13", "value": 67 }],
    "interest_by_region": [{ "geo_code": "US-HI", "region": "Hawaii", "value": 100 }],
    "related_queries": { "top": [{ "query": "coffee shop", "value": 100 }], "rising": [] }
}

Comparison (one item for the whole set — values share one scale):

{
    "type": "comparison",
    "keywords": ["netflix", "disney plus", "hbo max"],
    "scale": "shared",
    "stats": { "netflix": { "average": 59.1 }, "disney plus": { "average": 7.6 }, "hbo max": { "average": 5.3 } },
    "interest_over_time": [{ "date": "2026-04-17", "values": { "netflix": 61, "disney plus": 8, "hbo max": 5 } }],
    "interest_by_region": [{ "region": "California", "values": { "netflix": 58, "disney plus": 9, "hbo max": 6 } }],
    "related_queries": { "netflix": { "top": [], "rising": [] } }
}

Every run also pushes one summary item (what succeeded, what failed and why, what was charged), and failures push free error items with plain-language messages.

How much does it cost?

Event	Price	You pay when
Actor start	$0.02	each run starts
Trending Now	$0.02 per country	the country's trend list came back (up to ~480 trends + news — one flat price, however many items)
Keyword report	$0.03 per keyword	the report contains data
Comparison	$0.05 per comparison	the comparison contains data

Examples: daily "what's trending in the US" schedule ≈ $1.20/month. A 10-keyword research run = $0.32. Comparing 5 brands = $0.07.

Failed or empty results are never charged — that's the whole point.

FAQ

Where does the data come from? Google Trends' own public endpoints — the same ones the trends.google.com pages call. Nothing is estimated or modeled.

Can I use it from an AI agent / MCP? Yes — the input is flat and explicit, and every output item has a type field to route on. Ask your agent things like "compare interest in X vs Y over the last 3 months" or "what's trending in Germany right now".

Which countries work for Trending Now? Any country Google Trends supports (50+): US, GB, IN, DE, JP, BR… one geo code per entry in trendingGeos.

Why do keyword runs need the residential proxy? Google rate-limits Trends per IP aggressively. The default configuration (Apify residential) is what makes keyword reports reliable; datacenter proxies will get 429s. Trending Now doesn't need a proxy at all.

Is this legal? The Actor only reads publicly available data that Google publishes to everyone. It doesn't log in, bypass paywalls, or collect personal data.