Google Autocomplete Scraper - Keyword Suggestions

What is Google Autocomplete Scraper?

Google Autocomplete Scraper extracts keyword suggestions from the Google search box, the YouTube search box and the Google Shopping search box. Give it a seed keyword such as "running shoes" and it returns hundreds of real long-tail searches people type, each with Google's own relevance score, its position in the dropdown, and the exact query that surfaced it.

Autocomplete is Google telling you what people search for, in their own words, right now. It is the fastest free source of long-tail keywords, content ideas, FAQ questions and product research there is.

• 🔎 Hundreds of keywords from one seed. Alphabet expansion (a to z, 0 to 9) turns Google's 15 suggestions into 300 to 500 unique long-tail phrases
• ❓ Questions people actually ask, from how, what, why, can, is and seven more question words, ready for FAQ pages and blog outlines
• ⚖️ Comparisons and buyer intent from modifiers like "vs", "for", "near", "best" and "cheap"
• 📺 Three autocompletes in one actor: Google Search, YouTube and Google Shopping, each deduped on its own
• 📊 Google's relevance score on every row, so you can rank suggestions the way Google does instead of guessing
• 🌍 Any country and language, including non-Latin scripts such as Japanese, Arabic and Hindi
• 🔁 Recursive expansion that feeds each suggestion back in to dig deeper into a niche
• ⚡ Fast and cheap. Plain HTTP requests, no browser, no login, no Google API key. 500 keywords in about 10 seconds

What data does Google Autocomplete Scraper extract?

Every run writes one row per unique suggestion.

Field	Example
suggestion	how crm software works
keyword	crm software
source	google, youtube or shopping
relevance	1250 (Google's relevance score, higher is stronger)
position	1 (rank in the dropdown)
matchedQuery	how crm software (the query that surfaced it)
expansionType	seed, question, modifier, alphabet or recursive
expansionToken	how
depth	0 for direct expansions, 1 to 3 for recursive ones
wordCount / characterCount	4 / 22
isQuestion	true
containsKeyword	true when every word of the seed is in the suggestion
suggestionType / subtypes	QUERY / [512] (Google's internal type codes)
country / language	us / en
scrapedAt	2026-10-05T13:10:09.650Z

The dataset also has a ready Questions view that keeps only the question-shaped suggestions.

How to scrape Google autocomplete suggestions

1. Open Google Autocomplete Scraper and click Try for free.
2. Add one or more Seed keywords, for example "running shoes" or "crm software".
3. Pick the Autocomplete sources: Google Search, YouTube, Google Shopping, or all three.
4. Leave Alphabet expansion on for volume. Turn on Question expansion and Modifier expansion for content and buyer intent research.
5. Set the Country and Language of the market you are researching.
6. Click Start. Results appear live in the Output tab.
7. Export as JSON, CSV, Excel or XML, or pull them from the Apify API.

Input example

{
    "keywords": ["crm software"],
    "sources": ["google", "youtube"],
    "expandAlphabet": true,
    "expandQuestions": true,
    "expandModifiers": false,
    "depth": 0,
    "country": "us",
    "language": "en",
    "maxSuggestionsPerKeyword": 500
}

Output example

{
    "keyword": "crm software",
    "source": "google",
    "suggestion": "how crm software works",
    "position": 1,
    "relevance": 1250,
    "suggestionType": "QUERY",
    "subtypes": [512],
    "description": null,
    "matchedQuery": "how crm software",
    "expansionType": "question",
    "expansionToken": "how",
    "depth": 0,
    "wordCount": 4,
    "characterCount": 22,
    "isQuestion": true,
    "containsKeyword": true,
    "country": "us",
    "language": "en",
    "scrapedAt": "2026-10-05T13:10:09.650Z",
    "noResults": false,
    "notice": null
}

A few more real suggestions for "crm software": crm software for small businesses, crm software for real estate, which crm software is best, how much crm software cost, what crm software does verizon use.

How many keywords can I get?

Google shows up to 15 suggestions per query. With alphabet expansion one seed keyword typically yields 300 to 500 unique suggestions per source. Question and modifier expansion add up to about 150 more. Recursive depth 1 multiplies that roughly tenfold, and depth 2 or 3 goes further still, so set Max suggestions per keyword to keep big runs bounded.

Suggestions are deduped within each keyword and source, so you never pay twice for the same phrase from the same autocomplete.

How much does it cost to scrape Google autocomplete?

Google Autocomplete Scraper uses pay per result pricing: $0.40 per 1,000 keyword suggestions, with no monthly rental and no start fee. You only pay for unique suggestions written to your dataset.

Run	Rows	Cost
1 keyword, Google Search, alphabet expansion	about 500	about $0.20
10 keywords, Google Search, alphabet and questions	about 5,500	about $2.20
1 keyword, Google, YouTube and Shopping	about 1,500	about $0.60

Every Apify account includes free monthly usage credit, so you can try it at no cost. Set Max suggestions in total to cap what a run can spend.

Use cases

• SEO keyword research: build long-tail keyword lists and topic clusters around a seed term
• Content planning: pull the questions people ask for blog posts, FAQ pages and "People also ask" style sections
• YouTube SEO: find what viewers type into YouTube search before you title and tag a video
• Ecommerce and Amazon style product research: Google Shopping suggestions show the product attributes, sizes and brands buyers search for
• PPC and paid search: discover modifiers and comparisons ("vs", "cheap", "near me") for ad groups and negative keywords
• Brand monitoring: see what Google suggests next to your brand or a competitor's, in every market you sell in
• Trend spotting: schedule a daily run and diff the results to catch new searches as they appear

Integrations

Connect Google Autocomplete Scraper to Google Sheets, Airtable, Slack, Zapier, Make, n8n or your own code through Apify integrations, or call it from Python or JavaScript with the Apify API client. Schedule it to track how suggestions change over time.

FAQ

Is it legal to scrape Google autocomplete?

Google Autocomplete Scraper reads the same public suggestions anyone sees when typing into Google, YouTube or Google Shopping. It does not log in, use cookies or access anything behind an account, and suggestions contain no personal data. As with any data source, check that your use complies with the terms and laws that apply to you, and consult a lawyer if you are unsure.

How much does it cost?

$0.40 per 1,000 unique suggestions, billed per result. A keyword that returns nothing writes one notice row explaining why, billed as a single result ($0.0004).

Do I need a Google account, cookies or an API key?

No. The actor needs nothing from you but keywords. It runs on Apify Proxy, which is included with every Apify account.

What does the relevance score mean?

It is the score Google itself attaches to each suggestion when it builds the dropdown. The top suggestion usually scores 1250 or more, and the rest step down from about 600. Use it to rank suggestions within a query, or to spot the completions Google is most confident about.

Why do some suggestions not contain my keyword?

Google sometimes suggests a related phrasing, a singular form or a spelling correction ("running shoe" for "running shoes"). The containsKeyword field marks these so you can filter them out or keep them as related keywords.

Can I get suggestions for other countries and languages?

Yes. Set Country to a two letter code such as gb, de, in or br, and Language to the language of your keywords, such as de, es, ja or hi. Suggestions come back in their original script.

Does it include search volume?

No. Autocomplete shows what people search, not how often. Pair the output with a search volume tool, or use the relevance score and position as a free proxy for popularity.

Why did my run stop before the cap?

Each keyword has a natural ceiling: once every expansion query has been asked, there is nothing new to return. Turn on question and modifier expansion, or set Recursive depth to 1, to go deeper.