Google Autocomplete Scraper: Keyword Ideas from Google, YouTube & Shopping

Turn one keyword into hundreds of real long-tail search phrases. This Actor reads the suggestions Google shows while people type, from Google web search, YouTube, Google Shopping, Google Images and Google News, and expands every seed with A-Z, question and modifier variations. Any country, any language, in bulk.

It calls Google's suggestion service over plain HTTP (no browser), so it is fast and cheap: a typical seed with A-Z and question expansion finishes in a few seconds and returns 300-500 unique suggestions.

What you get

Feature	Details
Bulk seeds	As many keywords as you like in one run
5 sources	Google web, YouTube, Google Shopping, Images, News
A-Z expansion	keyword a ... keyword z
Question expansion	how keyword, what keyword, why keyword ...
Modifier expansion	keyword for, keyword with, keyword vs, keyword near ...
0-9 expansion	keyword 1 ... keyword 9
Depth	Get suggestions for every suggestion (depth 1-2) for really big lists
Relevance score	Google's own relevance number for each suggestion
Deduplication	Each suggestion appears once per seed and source
Filters	Max suggestions per seed; only suggestions that contain the seed
Any market	Country (gl) and language (hl) codes

Use cases

• SEO keyword research: long-tail phrases and questions for articles and FAQ sections.
• YouTube SEO: what people type into YouTube search, for titles, tags and video ideas.
• E-commerce and Amazon/Etsy sellers: Google Shopping suggestions show buyer-intent phrases.
• Book and product niches: see how people describe a niche in their own words.
• AI agents and automations: run it from n8n, Make, Zapier or the Apify MCP server.

Pair it with our Google Trends Scraper (ksvan/google-trends-scraper) to check which of the found phrases are rising.

Input example

{
  "keywords": ["cozy mystery", "air fryer"],
  "sources": ["web", "youtube"],
  "country": "us",
  "language": "en",
  "expandAlphabet": true,
  "expandQuestions": true,
  "depth": 0
}

Output example

{
  "keyword": "cozy mystery",
  "source": "web",
  "suggestion": "cozy mystery conventions",
  "depth": 0,
  "query": "cozy mystery c",
  "expansion": "alphabet",
  "position": 1,
  "relevance": 1000,
  "suggestionType": "QUERY",
  "wordCount": 3,
  "containsKeyword": true,
  "country": "us",
  "language": "en"
}

Pricing

Pay per result: you are charged for each unique suggestion saved to the dataset. Use Max suggestions per keyword or the run's maximum cost setting to control spend; the Actor stops cleanly when the limit is reached.

Tips

• Question expansion often returns loosely related ideas. Turn on Only suggestions containing the keyword for a tight list.
• Depth 1 multiplies the output (every suggestion is queried again). Start with depth 0.
• Suggestions differ by country and language; run separate seeds per market for clean data.

FAQ

Is this legal? The Actor collects only the public, anonymous suggestions that Google shows to everyone in its search boxes. No personal data is collected. Check your own use case and local laws.

Something broke or you need a feature? Open an issue on the Actor page.