Google Autocomplete Keyword Scraper

Pricing

Rating

Developer

Actor stats

Categories

Share

Harvest hundreds of real Google Autocomplete suggestions from a handful of seed keywords — no API key required.

Drop in your seeds, pick your expansion strategies, and get back the long-tail keywords people are actually typing into Google: questions, modifiers, comparisons, and "alphabet soup" variations. Perfect for SEO content planning, PPC keyword discovery, and finding the exact phrasing your audience uses.

⚡ Works out of the box — uses Google's public autocomplete endpoint over HTTP. No SerpApi key, no proxy needed for normal volumes.

What it does

For each seed keyword, the Actor sends a set of expanded queries to Google Autocomplete and collects every suggestion:

Output is one row per suggestion, tagged with the seed and the modifier that surfaced it, plus a deduplicated flat keyword list ready to paste into other tools.

Who it's for

• SEO & content teams — build topic clusters and find long-tail keywords fast.
• PPC marketers — discover real query phrasings for ad groups and negative keywords.
• YouTubers & creators — find the questions people ask about your niche.
• eCommerce sellers — surface product modifiers and comparison searches.
• Anyone doing keyword research who wants raw Google demand language without a paid SEO suite.

Popular keyword research use cases

• Long-tail SEO keyword research - expand seed terms into Google Autocomplete suggestions.
• Question keyword discovery - find "how", "what", "why", and "best" queries for content briefs.
• Local SEO content ideas - run city/service seeds to uncover real local search phrases.
• eCommerce product modifiers - discover product attributes, alternatives, comparisons, and buying-intent terms.
• YouTube and creator topics - turn niche ideas into video titles and search-led content angles.
• PPC ad group and negative keyword research - collect real phrasing before building campaigns.

Input

At least one of keywords / keywordsText is required.

Example input

{
"keywords":["ai seo tools","content marketing","email newsletter"],
"geo":"US",
"language":"en",
"expansions":["questions","prepositions","alphabet"]
}

More in examples/.

Output

One dataset row per suggestion:

{
"seed":"ai seo tools",
"query":"what ai seo tools",
"suggestion":"what is ai seo tools",
"modifier":"what",
"modifierType":"questions",
"position":1,
"geo":"US",
"language":"en",
"source":"google-autocomplete",
"fetchedAt":"2026-06-14T02:42:49.918Z"
}

Key-value store outputs

• KEYWORDS — deduplicated flat list of every suggestion (paste-ready)
• KEYWORDS_BY_TYPE — suggestions grouped by expansion type
• SUMMARY — totals, unique count, top seeds by volume
• ERRORS — any failed queries

How many keywords will I get?

Roughly seeds × queries-per-seed × ~5–10 suggestions, before dedup. One seed with questions + prepositions + alphabet sends ~51 queries and typically yields 100–300 raw suggestions. Scale up by adding seeds.

Request math also tells you cost: each query is one HTTP call. 100 seeds × 51 queries = ~5,100 calls — keep maxConcurrency modest and requestDelayMs ≥ 200 to stay clean, or enable a proxy.

Pricing

Pay-per-result — one suggestion row = one billable item. No external API cost in the default http mode.

Limitations & honesty

• Autocomplete reflects what Google suggests, not exact search volume. Pair with a volume tool if you need numbers.
• Suggestions are personalized/localized — results vary by geo and language.
• Very high volume from one IP can get rate-limited; raise requestDelayMs, lower maxConcurrency, or add a proxy.

Related tools

• Google Trends Keyword Monitor — rank keywords by rising/breakout trend.
• Google Trends Content Calendar Generator — turn trends into a dated publishing plan.

Changelog

See CHANGELOG.md.