Google Autocomplete Scraper: Keyword Suggestions

Google Autocomplete Scraper

Scrape Google autocomplete suggestions at scale for SEO keyword research, content planning, and PPC campaign expansion. Turn a handful of seed keywords into hundreds of long-tail keyword ideas using Google's own search predictions — in seconds.

🔍 What does Google Autocomplete Scraper do?

Google Autocomplete Scraper extracts keyword suggestions directly from Google's autocomplete API — the same predictions that appear when you start typing in the search bar. Enter seed keywords and receive every phrase Google suggests (also known as Google Suggest), with Google's own relevance score, organized by depth, language, and country.

The scraper communicates with Google's public suggest endpoint via lightweight HTTP requests — no browser required. This makes it significantly faster and cheaper than Playwright-based alternatives, and it runs comfortably on just 256 MB of memory.

Key capabilities:

• Extract up to 15 autocomplete suggestions per Google query — you pay per query, not per suggestion, so more suggestions cost nothing extra
• Google's relevance score and suggestion type on every result, so you can rank keyword ideas without a second tool
• Recursive expansion (depth 1–3): take each suggestion and fetch its own sub-suggestions to build deep keyword trees
• Alphabet append: query "keyword a", "keyword b", ... "keyword z" to unlock up to 27× more variations
• Many countries and languages in one run — set countries and languages, and every combination is queried
• Optional run-wide deduplication — keep each suggestion once and avoid paying to expand a repeat
• Keyword ideas only: website links and calculator answers from Google are filtered out, and the query itself is never returned as its own suggestion
• Exact query provenance and zero-based Google response position on every result
• Query-local deduplication: repeated values in one response are removed while overlaps across seeds and expansions stay attributable
• Export to JSON, CSV, Excel, or pipe results directly via API

Try it instantly: automation-lab/google-autocomplete-scraper — no API key or login required.

Who is it for? (Google Autocomplete Scraper users)

SEO specialists and content strategists

• Discover long-tail keywords with real user intent behind them
• Map the full keyword space around a topic before writing a content brief
• Identify questions and modifiers your competitors may have missed
• Track how autocomplete suggestions evolve as topics gain or lose momentum

PPC managers and growth marketers

• Expand Google Ads keyword lists with real search queries, not guesses
• Find negative keyword candidates by seeing what Google associates with your terms
• Research competitor brand terms and adjacent queries for targeting
• Build keyword clusters for ad group structuring

Market researchers and product teams

• Understand what consumers are actively searching for in your niche
• Identify emerging product categories or pain points before they hit mainstream tools
• Compare search behavior across countries and languages side by side
• Validate product names or feature ideas against actual search language

Affiliate marketers and niche researchers

• Find buyer-intent keyword variations for review and comparison content
• Discover low-competition long-tail keywords for new niche sites
• Research sub-niches and topic clusters for content roadmaps
• Run alphabet append on a niche term to exhaust all keyword angles quickly

✅ Why use Google Autocomplete Scraper?

• Real search behavior, not historical estimates — autocomplete reflects what users are typing right now, not what they searched last month
• Pure HTTP — no browser overhead — runs in under a second per keyword; no Playwright, no Chromium, no wasted compute
• Recursive expansion built in — go 1–3 levels deep to build keyword trees automatically
• No API key or Google account needed — works out of the box with zero setup
• Fully localized — get suggestions for any language and country combination; discover local intent unavailable in English-only tools
• Alphabet append for maximum coverage — one click to query all 26 letter suffixes, uncovering suggestions most SEO tools miss
• Apify platform advantages — scheduling, proxy rotation, API access, export to Google Sheets, Zapier, Make, and 5,000+ apps
• Pay only for what you use — no monthly subscription; run-start and per-keyword charges are tiered by Apify plan

📊 What data can you extract?

Field	Type	Description
keyword	string	The original seed keyword you provided
query	string	The exact direct, alphabet, prefix, or recursive query sent to Google
suggestion	string	The autocomplete suggestion returned by Google
position	integer	Zero-based order in Google's response. Numbers can skip where a website link or calculator answer was left out
relevance	integer	Google's own relevance score for this suggestion. Higher ranks higher
suggestionType	string	Google's type for this suggestion, usually QUERY
depth	integer	Expansion depth: 0 = direct suggestion, 1 = sub-suggestion, 2 = deeper
parentKeyword	string	The query one level above query: the seed for alphabet and prefix queries, the earlier query for deeper levels, and null for the seed's own direct query
language	string	Language code used for the query (e.g., en, es, de)
country	string	Country code used for the query (e.g., us, gb, jp)
scrapedAt	string	ISO 8601 timestamp of when the suggestion was extracted

Result counts by configuration:

Configuration	Approximate results per seed keyword
Depth 1 (default)	Up to 10 suggestions (15 with maxSuggestionsPerKeyword: 15)
Depth 2	Up to 110 suggestions (10 + 10×10)
Depth 3	Up to 1,110 suggestions
Depth 1 + alphabet append	Up to 270 suggestions (27 × 10)
Depth 2 + alphabet append	Up to 2,970 suggestions

💰 How much does it cost to scrape Google autocomplete suggestions?

This actor uses pay-per-event pricing — you pay only for what you scrape. No monthly subscription. All platform costs are included.

Apify plan	Run start	Per keyword expanded
Free	$0.005	$0.001
Bronze	$0.00475	$0.0009
Silver	$0.00425	$0.0008
Gold	$0.00375	$0.00065
Platinum	$0.003	$0.0005
Diamond	$0.0025	$0.0004

From November 23, 2026 the price will be: run start at today's Bronze price on every plan; per keyword expanded at today's Bronze price on Free and Bronze, 10% below it on Silver, and 20% below it on Gold, Platinum and Diamond.

Calculated cost examples (not event unit prices)

Task	Seed keywords	Queries fired	Cost (Free tier)
Quick idea check	1 keyword, depth 1	1	~$0.006
Standard research session	5 keywords, depth 1	5	~$0.010
Deep keyword tree	1 keyword, depth 2	up to 11	~$0.016
Alphabet sweep	1 keyword, depth 1 + alphabet	27	~$0.032
Full niche map	5 keywords, depth 2 + alphabet	up to 1,485	~$1.49

You pay per Google query, not per suggestion. One query returns up to 15 suggestions, which works out to about $0.0001 per suggestion or less. A query that returns nothing, a query that fails, and a run with invalid input are not charged.

Free plan value: With Apify's $5 monthly credit, you can run approximately 4,900 keyword queries — enough to build a thorough keyword map for a new niche or content site.

🚀 How to scrape Google autocomplete suggestions

1. Go to Google Autocomplete Scraper on the Apify Store
2. Click Try for free — no account required for the first run
3. In the Keywords field, enter one or more seed keywords (e.g., web scraping, python tutorial)
4. Set Language and Country for localized results (default: English / United States)
5. Choose Expansion Depth: 1 for a quick look, 2 for a comprehensive keyword list, 3 for an exhaustive tree
6. Optionally enable Append Alphabet to query all 26 letter suffixes for maximum coverage
7. Click Save & Start — the run typically completes in under 30 seconds for small inputs
8. When finished, click Export to download your suggestions as JSON, CSV, or Excel

Input example — quick keyword idea check:

{
    "keywords": ["content marketing"],
    "language": "en",
    "country": "us",
    "maxDepth": 1,
    "appendAlphabet": false
}

Input example — full niche keyword map:

{
    "keywords": ["affiliate marketing", "passive income ideas"],
    "language": "en",
    "country": "us",
    "maxDepth": 2,
    "appendAlphabet": true,
    "maxSuggestionsPerKeyword": 10
}

Input example — question and numeric intent discovery:

{
    "keywords": ["running shoes"],
    "expansionPatterns": ["questionPrefixes", "numbers"],
    "maxDepth": 1
}

This opt-in mode keeps the standard direct query and adds bounded prefix queries such as how running shoes and 5 running shoes. Inputs that omit expansionPatterns behave exactly as before.

Input example — international SEO research (Spanish, Mexico):

{
    "keywords": ["marketing digital", "herramientas seo"],
    "language": "es",
    "country": "mx",
    "maxDepth": 1,
    "appendAlphabet": false
}

⚙️ Input parameters

Parameter	Type	Default	Description
keywords	array	required	One or more seed keywords to expand. Each becomes a root query.
language	string	"en"	ISO 639-1 language code for suggestions. Examples: en, es, de, fr, ja, pt-BR, zh.
country	string	"us"	Two-letter ISO 3166-1 country code. Examples: us, gb, de, fr, jp, br. UK is treated as gb. A code that is not two letters, such as USA, stops the run with a clear message and no charge.
languages	string[]	—	Optional. Several languages in one run. Overrides language. Each language is combined with each country.
countries	string[]	—	Optional. Several countries in one run. Overrides country. Each country is combined with each language.
maxDepth	integer	1	How many levels deep to recursively expand suggestions. 1 = direct suggestions only. 2 = also expand each direct suggestion. 3 = one more level. Max: 3.
maxSuggestionsPerKeyword	integer	10	Maximum suggestions to save per query, 1 to 15. Google returns up to 15, and the price per query stays the same.
appendAlphabet	boolean	false	Also query "keyword a", "keyword b", ... "keyword z" to find additional variations. Multiplies query count by ~27× but significantly increases coverage.
expansionPatterns	string[]	[]	Optional extra fan-out before each seed: questionPrefixes queries who/what/when/where/why/how/can/is/are/should, while numbers queries 0-9. Each added query uses one keyword-expanded event.
deduplicateSuggestions	boolean	false	Save each suggestion once per language and country across the run, and do not expand a repeat. Off by default, so every query that produced a suggestion stays visible.
maxRequestRetries	integer	3	Maximum retry attempts for failed HTTP requests. Increase if you see timeouts.

📄 Output examples

Depth 0 result (direct suggestion from seed keyword):

{
    "keyword": "web scraping",
    "query": "web scraping",
    "suggestion": "web scraping python",
    "position": 0,
    "relevance": 601,
    "suggestionType": "QUERY",
    "depth": 0,
    "parentKeyword": null,
    "language": "en",
    "country": "us",
    "scrapedAt": "2026-03-23T12:00:00.000Z"
}

Depth 1 result (sub-suggestion — expanded from a depth-0 suggestion):

{
    "keyword": "web scraping",
    "query": "web scraping python",
    "suggestion": "web scraping python tutorial for beginners",
    "position": 2,
    "relevance": 521,
    "suggestionType": "QUERY",
    "depth": 1,
    "parentKeyword": "web scraping",
    "language": "en",
    "country": "us",
    "scrapedAt": "2026-03-23T12:00:05.000Z"
}

International result (Spanish, Spain):

{
    "keyword": "marketing digital",
    "query": "marketing digital",
    "suggestion": "marketing digital para principiantes",
    "position": 0,
    "relevance": 601,
    "suggestionType": "QUERY",
    "depth": 0,
    "parentKeyword": null,
    "language": "es",
    "country": "es",
    "scrapedAt": "2026-03-23T12:01:00.000Z"
}

💡 Tips for best results

• Start with depth 1 to quickly evaluate what Google associates with your seed keywords, then increase depth for long-tail research
• Use alphabet append for exhaustive coverage — it exposes suggestions that only appear when a letter suffix is added, often revealing the most specific long-tail phrases
• Combine depth 2 + alphabet on a single focused seed keyword for the most thorough keyword map of one topic (this produces 2,970+ suggestions)
• Use specific seed keywords — "best running shoes for flat feet" yields more actionable suggestions than "shoes"; the more specific, the more relevant the sub-suggestions
• Try the same keyword in multiple languages to discover international keyword gaps — suggestions often differ substantially between en/us and en/gb
• Run on a schedule (weekly or monthly) to track how autocomplete evolves as topics trend or fade — Apify Console makes it easy to set recurring runs
• Filter depth 0 results first — these are the most popular direct associations; use them to prioritize which branches to expand
• Keep maxDepth at 1 for PPC campaigns — depth-1 suggestions are the most commercially relevant and least noisy

🔗 Integrations

Google Autocomplete Scraper connects to your existing workflow tools via the Apify platform:

Google Autocomplete Scraper → Google Sheets Set up a scheduled run to automatically push fresh keyword suggestions into a Google Sheets tab weekly. Use the built-in Google Sheets integration in Apify Console — no code needed. Your SEO team always has up-to-date keyword ideas without manual research.

Google Autocomplete Scraper → Ahrefs or SEMrush Export the scraper's output as CSV and import it into Ahrefs' Keywords Explorer or SEMrush's Keyword Magic Tool. Use autocomplete-discovered keywords as the seed list for volume and difficulty analysis — this hybrid approach finds more keywords than either tool would alone.

Google Autocomplete Scraper → Google Ads Editor Run the scraper with appendAlphabet: true on your core product terms to generate a comprehensive long-tail keyword list. Export as CSV and import directly into Google Ads Editor for campaign keyword expansion without manual brainstorming.

Google Autocomplete Scraper → Zapier / Make Trigger the scraper via webhook whenever a new product or topic is added to your system. Route results automatically to Slack, Airtable, Notion, or your content management system. Build a "keyword on demand" pipeline that fires without manual intervention.

Google Autocomplete Scraper → Slack alerts Schedule weekly keyword research runs and use Apify's webhook integration to post a summary of newly discovered suggestions to your team's Slack channel. Keep the whole team aware of emerging search trends in your niche.

Google Autocomplete Scraper → Content calendars Feed keyword trees (depth 2) into your content planning tool via API. Each top-level suggestion becomes a pillar page topic; each sub-suggestion becomes a supporting cluster article. Build a data-driven content calendar in minutes.

API usage

Run Google Autocomplete Scraper programmatically from your own applications or scripts.

Node.js

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_API_TOKEN' });

const run = await client.actor('automation-lab/google-autocomplete-scraper').call({
    keywords: ['content marketing', 'email marketing'],
    language: 'en',
    country: 'us',
    maxDepth: 2,
    appendAlphabet: false,
    maxSuggestionsPerKeyword: 10,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(`Found ${items.length} keyword suggestions`);
items.forEach(item => {
    console.log(`[${item.query} #${item.position}] ${item.suggestion}`);
});

Python

from apify_client import ApifyClient

client = ApifyClient('YOUR_API_TOKEN')

run = client.actor('automation-lab/google-autocomplete-scraper').call(run_input={
    'keywords': ['content marketing', 'email marketing'],
    'language': 'en',
    'country': 'us',
    'maxDepth': 2,
    'appendAlphabet': False,
    'maxSuggestionsPerKeyword': 10,
})

items = client.dataset(run['defaultDatasetId']).list_items().items
print(f"Found {len(items)} keyword suggestions")
for item in items:
    print(f"[{item['query']} #{item['position']}] {item['suggestion']}")

cURL

curl -X POST \
  "https://api.apify.com/v2/acts/automation-lab~google-autocomplete-scraper/runs?token=YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "keywords": ["content marketing"],
    "language": "en",
    "country": "us",
    "maxDepth": 1,
    "appendAlphabet": false
  }'

To retrieve the results after the run completes:

$curl "https://api.apify.com/v2/datasets/RUN_DEFAULT_DATASET_ID/items?token=YOUR_API_TOKEN"

🤖 Use with AI agents via MCP

Google Autocomplete Scraper is available as a tool for AI assistants that support the Model Context Protocol (MCP).

Add the Apify MCP server to your AI client — this gives you access to all Apify actors, including this one:

Setup for Claude Code

$claude mcp add --transport http apify "https://mcp.apify.com?tools=automation-lab/google-autocomplete-scraper"

Setup for Claude Desktop, Cursor, or VS Code

Add this to your MCP config file:

{
    "mcpServers": {
        "apify": {
            "url": "https://mcp.apify.com?tools=automation-lab/google-autocomplete-scraper"
        }
    }
}

Your AI assistant will use OAuth to authenticate with your Apify account on first use.

Example prompts

Once connected, try asking your AI assistant:

• "Use automation-lab/google-autocomplete-scraper to find all keyword suggestions for 'project management software' in English with depth 2, then organize the results by topic cluster"
• "Get Google autocomplete suggestions for 'keto diet' and 'intermittent fasting' and identify which long-tail keywords appear for both — those are the highest-opportunity topics"
• "Run the Google autocomplete scraper with alphabet append on 'email marketing' to build a complete keyword map, then highlight the top 20 most specific long-tail suggestions"

Learn more in the Apify MCP documentation.

⚖️ Is it legal to scrape Google autocomplete suggestions?

Google Autocomplete Scraper accesses Google's public suggest API — the same endpoint your browser calls when you type in the search bar. No login, no account, no authentication is required. The suggestions returned are freely provided to every internet user.

Scraping publicly available data is generally considered lawful. The US Ninth Circuit Court of Appeals (hiQ Labs v. LinkedIn, 2022) affirmed that accessing publicly available web data does not violate the Computer Fraud and Abuse Act. Similar legal precedents apply in the EU and UK.

Best practices:

• Use autocomplete data only for legitimate keyword research, SEO analysis, and market research purposes
• Do not use suggestions to build spam databases or engage in manipulative search behavior
• Review Google's Terms of Service for the latest policy updates
• For datasets involving personal data, ensure compliance with GDPR, CCPA, and other applicable privacy regulations

For more on the legality of web scraping, see Apify's web scraping legality guide.

❓ FAQ

How many suggestions does Google return per keyword? Google returns up to 15 autocomplete suggestions per query, and this Actor saves 10 by default (set maxSuggestionsPerKeyword to 15 for all of them; the price stays the same). The actual number depends on the keyword and locale — very new or niche terms may return fewer. Use alphabet append or increase depth to find more variations for low-suggestion keywords.

How fast does the scraper run? Very fast. Each keyword query is a simple HTTP request that resolves in under 200ms. A standard run with 5 seed keywords at depth 1 completes in under 5 seconds. Even a large alphabet-append run (270 queries) typically finishes in under 2 minutes.

How does this compare to paid keyword tools like Ahrefs or SEMrush? Those tools show search volume and keyword difficulty, but their keyword databases are updated infrequently and miss the freshest long-tail terms. Google Autocomplete shows real-time search behavior that keyword tools haven't yet indexed. The best approach is to use both: use this scraper to discover keyword ideas, then import them into Ahrefs or SEMrush for volume data.

Why am I getting fewer results than expected? Some keywords have fewer suggestions in Google's system — this is normal for very niche, misspelled, or brand-new terms. If you expect more results, try: (1) enabling alphabet append, (2) increasing maxDepth to 2, or (3) trying slight variations of your seed keyword. Also verify your language/country combination is correct — language: "es" with country: "us" returns Hispanic-US suggestions, which may differ from country: "es".

Will I see the seed keyword itself as a suggestion, or website links? No. The query itself, website links, and calculator answers are left out, because they are not keyword ideas and expanding them would waste paid queries.

What does a run cost when nothing comes back? Nothing. The run start and query charges only apply once a query delivers suggestions. Queries that return no suggestions or fail after retries are never charged, and invalid input stops before any charge.

What if requests fail or return errors? The scraper retries failed requests automatically with a rising delay. If you see failures despite retries, increase maxRequestRetries to 5–10. The run's final status tells you how many suggestions were saved, how many queries answered, and how many failed.

Can I run the scraper on a schedule to track keyword trends over time? Yes. Use Apify Console's built-in scheduler to run the actor daily, weekly, or monthly with the same input. Each run stores results in a separate dataset. You can compare datasets over time to see which suggestions appear, disappear, or change in phrasing — a useful signal for emerging topics.

🔗 Other Google and SEO scrapers

• Google Search Scraper — scrape organic SERP results, People Also Ask, and related searches from Google Search
• Google Shopping Scraper — extract product listings, prices, and merchant data from Google Shopping
• Google News Scraper — scrape Google News headlines, articles, and topic clusters
• Google Scholar Scraper — extract academic papers, citations, and author profiles from Google Scholar
• Google Trends Scraper — track keyword popularity trends and breakout searches over time
• Google Maps Scraper — extract business listings, reviews, and contact info from Google Maps
• DuckDuckGo Scraper — scrape search results from DuckDuckGo for privacy-focused SERP research
• Keyword Density Analyzer — analyze keyword frequency and density across any web page for on-page SEO audits

🔐 Diagnostics and privacy

When a run fails, this Actor sends sanitized diagnostic input, the error, and the Actor, build and run IDs to our private GlitchTip service so we can repair it. Secret fields and URL query strings are removed, and reports are kept for 30 days.

Server API: get one keyword's suggestions over HTTP

Google Autocomplete Scraper is also a Server Actor. Besides normal runs, it answers short HTTP requests directly, without starting a run for each keyword. Copy the server URL from the Actor's Server tab (do not build it from the Actor name) and authenticate with your Apify token. Use normal Actor runs for many keywords, several languages or countries, and deeper expansion (maxDepth above 1).

curl -sS "$SERVER_URL/suggestions?keyword=coffee%20maker" -H "Authorization: Bearer $APIFY_TOKEN"
curl -sS "$SERVER_URL/suggestions?keyword=kaffee&language=de&country=de&maxSuggestions=5" -H "Authorization: Bearer $APIFY_TOKEN"
curl -sS "$SERVER_URL/suggestions?keyword=running%20shoes&appendAlphabet=true&expansionPattern=questionPrefixes&deduplicate=true" -H "Authorization: Bearer $APIFY_TOKEN"

Method and path	Input	Output and charge
GET /	None	ready: true, the routes and limits; no scraping or charge
GET /suggestions	Query keyword: one seed keyword (up to 200 characters); optional language (such as en, de or pt-BR; default en), country (two letters such as us or gb; default us), maxSuggestions (1–15; default 10), appendAlphabet (true adds "keyword a" to "keyword z"), expansionPattern (repeat for questionPrefixes and numbers), deduplicate (true returns each suggestion once)	{ok,count,results} with one row per suggestion, the same record as a normal dataset item at depth 0; one keyword-expanded event per query that returned suggestions (1 for the keyword alone, up to 47 with every expansion); no start fee

A keyword with no suggestions returns an error and is free. Each request has a 45-second deadline and a 1 MiB response cap, and one server works on at most eight requests at once. Nothing is cached. Errors use { "ok": false, "error": { "code": "...", "message": "...", "retryable": false } }: 400 INVALID_INPUT; 402 spending limit; 404 NO_SUGGESTIONS or unknown route; 405 wrong method; 413 RESPONSE_TOO_LARGE; 502 UPSTREAM_ERROR or charging failure; 503 server busy or stopping; 504 TIMEOUT. Failed requests are free. Retry only when retryable is true. The Server tab shows the full OpenAPI description.