Keyword Suggestions – Google, YouTube & Amazon Autocomplete

Keyword suggestions from Google, YouTube, Amazon, Bing and DuckDuckGo autocomplete in one run. Expand a seed with A–Z, numbers, questions, prepositions or your own modifiers, up to two levels deep, for any country and language. One row per suggestion, no start fee.

Price: $0.50 per 1,000 keyword suggestions. No start fee.

What does the Keyword Suggestions Scraper do?

It sends your keywords to the same public autocomplete services that power each site's search box and saves every suggestion with its context:

• Google Search autocomplete, up to 15 suggestions per query
• YouTube search suggestions, for video titles and topic research
• Bing autocomplete, up to 20+ suggestions per query
• DuckDuckGo autocomplete
• Amazon search suggestions for 21 marketplaces (US, CA, MX, BR, UK, DE, FR, IT, ES, NL, SE, PL, BE, TR, IN, AE, SA, EG, JP, AU, SG)

Expand mode turns one seed into hundreds of long-tail keywords:

Expansion	Queries sent for seed "coffee maker"
alphabet	coffee maker a, coffee maker b … coffee maker z
numbers	coffee maker 0 … coffee maker 9
questions	how / what / why / can / where / when / which / who / is / are / does / will coffee maker
prepositions	coffee maker vs / for / near / with / without / to / like / or / and / in / versus / is
custom	your own words, e.g. best {keyword} 2026, günstig
depth 2	every suggestion found is queried again ("suggestions of suggestions")

Duplicates are removed automatically. Suggestions identical to the seed or to the query sent are skipped, so you never pay for them.

Use cases

• SEO keyword research. Find long-tail keywords, questions for FAQ sections and content ideas.
• YouTube SEO. See what viewers actually type into YouTube search.
• Amazon and e-commerce. Find buyer-intent keywords for listings, PPC campaigns and product research.
• PPC and Google Ads. Build seed lists and negative-keyword candidates.
• Multi-country and multilingual research. Compare suggestions across countries and languages in one dataset.
• AI agents and content pipelines. Feed real search demand into LLM-generated briefs and outlines.

How to use it

1. Click Try for free.
2. Enter one or more Seed keywords.
3. Pick Search engines, your Country (us, gb, de, …) and Language (en, de, …).
4. Optional: switch Mode to expand, choose expansions and set Depth to 2.
5. Click Start, then download the results as CSV, Excel or JSON, or use the API.

Input example

{
    "keywords": ["coffee maker", "seo tools"],
    "engines": ["google", "youtube", "bing", "duckduckgo", "amazon"],
    "mode": "expand",
    "expansions": ["alphabet", "questions", "prepositions"],
    "customModifiers": ["best {keyword} 2026"],
    "depth": 1,
    "country": "us",
    "language": "en",
    "maxResultsPerKeyword": 500
}

Running with no input ({}) returns the suggestions for "coffee maker" from all five engines (about 60 rows).

Output

One row per suggestion. This is real output:

[
    { "type": "suggestion", "seed": "coffee maker", "suggestion": "coffee maker with grinder", "engine": "google", "modifier": null, "modifierType": null, "position": 1, "country": "us", "language": "en", "depth": 1, "query": "coffee maker", "parent": null },
    { "type": "suggestion", "seed": "coffee maker", "suggestion": "ninja coffee maker", "engine": "amazon", "modifier": null, "modifierType": null, "position": 2, "country": "us", "language": "en", "depth": 1, "query": "coffee maker", "parent": null },
    { "type": "suggestion", "seed": "coffee maker", "suggestion": "how coffee makers work", "engine": "youtube", "modifier": "how", "modifierType": "question", "position": 1, "country": "us", "language": "en", "depth": 1, "query": "how coffee maker", "parent": null },
    { "type": "suggestion", "seed": "coffee maker", "suggestion": "coffee maker vs espresso machine", "engine": "google", "modifier": "vs", "modifierType": "preposition", "position": 1, "country": "us", "language": "en", "depth": 1, "query": "coffee maker vs", "parent": null },
    { "type": "suggestion", "seed": "coffee maker", "suggestion": "coffee maker vs coffee machine", "engine": "bing", "modifier": "vs", "modifierType": "preposition", "position": 1, "country": "us", "language": "en", "depth": 1, "query": "coffee maker vs", "parent": null },
    { "type": "suggestion", "seed": "kahve makinesi", "suggestion": "kahve makinesi fiyatları", "engine": "google", "modifier": null, "modifierType": null, "position": 1, "country": "tr", "language": "tr", "depth": 1, "query": "kahve makinesi", "parent": null }
]

Field	Meaning
seed	Your input keyword
suggestion	The autocomplete suggestion
engine	google, youtube, bing, duckduckgo or amazon
modifier / modifierType	Modifier used (a, how, vs, …) and its group (alphabet, number, question, preposition, custom, suggestion). null for the plain seed.
position	1-based rank in that engine's list
country, language	Market used
depth, query, parent	Depth (1 or 2), the exact text sent, and for depth 2 the suggestion that was expanded

Free status rows explain anything unusual: ENGINE_UNSUPPORTED_COUNTRY (for example Amazon in a country without a marketplace), ENGINE_BLOCKED (an engine kept rate-limiting), NO_SUGGESTIONS and NO_RESULTS.

By default a suggestion appears once per engine, so you can see which engines agree. Turn on Remove duplicates across engines to keep each suggestion only once per seed.

Pricing

Pay-per-event: you only pay for suggestions saved, and platform usage is included. There is no start fee.

Event	Price	When it is charged
keyword	$0.0005 per suggestion ($0.50 / 1,000)	Each suggestion row saved
Status rows	Free	Unsupported country, blocked engine, no results

Job	Approx. rows	Cost
1 seed, simple, 5 engines	~60	$0.03
1 seed, expand (A–Z + questions + prepositions), 5 engines	~1,000–2,000	~$0.50–1.00
100 seeds, simple, Google only	~1,400	~$0.70

Use Max suggestions per seed keyword and Apify's Max cost per run to cap spending. The Actor stops exactly at your budget.

Use with AI agents (MCP)

This Actor works as a tool for AI agents through the Apify MCP server (Claude, Cursor, VS Code, OpenAI Agents and others). Example tool input:

{ "keywords": ["standing desk"], "engines": ["google", "amazon"], "mode": "expand", "expansions": ["questions"], "country": "gb", "language": "en", "maxResultsPerKeyword": 100 }

The agent gets flat rows (seed, suggestion, engine, modifier, position) that it can group, cluster or turn into content briefs directly.

FAQ

Does it return search volume? No. Autocomplete services don't publish volumes. They show what people actually type, ordered by popularity, so position is a useful relative signal. Combine the output with a volume tool if you need exact numbers.

How do country and language work? country selects the Amazon marketplace, the Bing market and the DuckDuckGo region, and is passed to Google and YouTube as gl. language is passed as hl and used for Bing and DuckDuckGo. Google also personalizes by the location of the request IP. For the most local Google results, enable Apify Proxy with a proxy country matching your target country.

Are the modifiers translated? The built-in question and preposition words are English. For other languages, add your own words in Custom modifiers, for example wie {keyword} or günstig. The alphabet and number expansions work in any language.

Will I get blocked? The Actor sends one request at a time per engine with a polite delay (default 250 ms) and retries with backoff. In testing, 915 requests (3 seeds × 61 queries × 5 engines) finished in about a minute with zero failures. If an engine still rate-limits a very large run, it is disabled with a free ENGINE_BLOCKED row and the other engines continue.

Is Amazon really included? Yes. The Actor uses Amazon's public search-completion endpoint, the one the amazon.* search box uses. It was verified for all 21 listed marketplaces.

Is it legal?

The Actor only reads public autocomplete suggestions, the same text anyone sees while typing in a search box. It does not log in and does not collect personal data. Suggestions are generated by each search engine and may occasionally contain brand names or trademarks. Use the data in line with each site's terms and applicable law.

Related actors

Other actors by the same developer:

• Google Trends Scraper – Interest Over Time & Related Queries – check how demand for the keywords you found changes over time
• YouTube Transcript Scraper – Captions, Subtitles & Video Data – read what the top videos for a keyword actually say
• Website SEO Audit Crawler – Broken Links, Schema & llms.txt – audit the pages you plan to target with these keywords

Support & feedback

Missing an engine or a country? Open an issue in the Issues tab.