Serper
App · Web Search · 8 actions
Run Google searches from a workflow and get the results as data: web results, images, news, videos, places, shopping, Scholar papers, and search suggestions.
Actions
| Action | What it does |
|---|---|
| Get search suggestions | Get the search suggestions Google would offer for a partial query, the way its search box does. |
| Search Google | Run a Google search and get the organic results back as data, along with the knowledge panel, answer box, 'People also ask' questions, and related searches when Google shows them. |
| Search Google Images | Search Google Images and get each picture's full-size and thumbnail address, its dimensions, and the page it came from. |
| Search Google Maps places | Find local businesses and places on Google Maps, with their address, category, phone number, website, rating, and coordinates. |
| Search Google News | Search Google News and get the matching articles with their headline, publisher, publication date, and link. |
| Search Google Scholar | Search academic literature on Google Scholar and get each paper's publication details, citation count, and PDF link where one is available. |
| Search Google Shopping | Search Google Shopping and get the matching products with their price, seller, delivery terms, and rating. |
| Search Google Videos | Search Google Videos and get the matching videos with their title, channel, length, and link. |
Get search suggestions
Get the search suggestions Google would offer for a partial query, the way its search box does. Useful for keyword research and for expanding a topic into the phrasing people actually search for.
Inputs
- Search query
querytextRequired - The partial query to get suggestions for, such as 'best coffee grinder for'.
- Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
Outputs
credits_usednumber- How many Serper credits this search spent.
suggestion_countnumber- How many suggestions were returned.
suggestionslist- The suggested searches, most likely first.
Example output
{
"credits_used": 1,
"suggestion_count": 2,
"suggestions": [
"best coffee grinder for espresso",
"best coffee grinder for french press"
]
}Search Google
Run a Google search and get the organic results back as data, along with the knowledge panel, answer box, 'People also ask' questions, and related searches when Google shows them.
Inputs
- Search query
querytextRequired - What to search for. Google's own operators work here, such as site:example.com or "an exact phrase".
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
- Location
locationtext - Search from a specific place, which changes the local results Google shows. Write it the way Google does, most specific part first.
- Published within
date_rangetext - Only return results published recently. Leave blank for any time.One of:
qdr:h(Past hour),qdr:d(Past 24 hours),qdr:w(Past week),qdr:m(Past month),qdr:y(Past year). - Correct spelling
autocorrecttrue/false - Let Google fix an apparent typo in the query. Turn this off when the query is a product code, a surname, or anything else Google would 'correct' into the wrong thing.Default:
true.
Outputs
answer_boxobject- The direct answer Google showed above the results, or null if there was none.
credits_usednumber- How many Serper credits this search spent.
knowledge_graphobject- The knowledge panel Google showed beside the results, or null if there was none.
people_also_asklist- The 'People also ask' questions and their snippets.
related_searcheslist- Related searches Google suggested.
result_countnumber- How many organic results were returned.
resultslist- The organic results, each with its position, title, link, snippet, date, and sitelinks.
Example output
{
"answer_box": null,
"credits_used": 1,
"knowledge_graph": {
"attributes": null,
"description": "Coffee is a beverage brewed from roasted coffee beans.",
"description_link": "https://en.wikipedia.org/wiki/Coffee",
"description_source": "Wikipedia",
"image_url": null,
"title": "Coffee",
"type": "Beverage",
"website": null
},
"people_also_ask": [
{
"link": "https://example.com/coffee-health",
"question": "Is coffee good for you?",
"snippet": "In moderation, coffee is fine for most adults.",
"title": "Coffee and health"
}
],
"related_searches": [
"coffee beans",
"types of coffee"
],
"result_count": 1,
"results": [
{
"attributes": null,
"date": null,
"link": "https://en.wikipedia.org/wiki/Coffee",
"position": 1,
"sitelinks": [],
"snippet": "Coffee is a beverage brewed from roasted coffee beans.",
"title": "Coffee - Wikipedia"
}
]
}Search Google Images
Search Google Images and get each picture's full-size and thumbnail address, its dimensions, and the page it came from.
Inputs
- Search query
querytextRequired - What kind of image to search for.
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
- Location
locationtext - Search from a specific place, which changes the local results Google shows. Write it the way Google does, most specific part first.
- Published within
date_rangetext - Only return results published recently. Leave blank for any time.One of:
qdr:h(Past hour),qdr:d(Past 24 hours),qdr:w(Past week),qdr:m(Past month),qdr:y(Past year).
Outputs
credits_usednumber- How many Serper credits this search spent.
image_countnumber- How many images were returned.
imageslist- The images, each with its title, full-size and thumbnail address, size, source, and page link.
Example output
{
"credits_used": 1,
"image_count": 1,
"images": [
{
"image_height": 1080,
"image_url": "https://example.com/coffee.jpg",
"image_width": 1920,
"link": "https://example.com/coffee",
"position": 1,
"source": "Example",
"thumbnail_url": "https://example.com/coffee-thumb.jpg",
"title": "A cup of coffee"
}
]
}Search Google Maps places
Find local businesses and places on Google Maps, with their address, category, phone number, website, rating, and coordinates.
Inputs
- Search query
querytextRequired - What kind of place to look for, such as 'coffee shop' or 'emergency dentist'.
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
- Location
locationtext - Search from a specific place, which changes the local results Google shows. Write it the way Google does, most specific part first.
Outputs
credits_usednumber- How many Serper credits this search spent.
place_countnumber- How many places were returned.
placeslist- The places, each with its title, address, category, phone number, website, rating, review count, and coordinates.
Example output
{
"credits_used": 1,
"place_count": 1,
"places": [
{
"address": "401 Congress Ave #100c, Austin, TX 78701",
"category": "Coffee shop",
"cid": "11265938073076301333",
"latitude": 30.266216,
"longitude": -97.743065,
"phone_number": "+1 512-394-6051",
"position": 1,
"rating": 4.5,
"rating_count": 1200,
"title": "Houndstooth Coffee",
"website": "https://example.com"
}
]
}Search Google News
Search Google News and get the matching articles with their headline, publisher, publication date, and link.
Inputs
- Search query
querytextRequired - What to look for in the news.
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
- Location
locationtext - Search from a specific place, which changes the local results Google shows. Write it the way Google does, most specific part first.
- Published within
date_rangetext - Only return results published recently. Leave blank for any time.One of:
qdr:h(Past hour),qdr:d(Past 24 hours),qdr:w(Past week),qdr:m(Past month),qdr:y(Past year).
Outputs
article_countnumber- How many articles were returned.
articleslist- The news articles, each with its title, link, snippet, date, source, and image.
credits_usednumber- How many Serper credits this search spent.
Example output
{
"article_count": 1,
"articles": [
{
"date": "2 hours ago",
"image_url": "https://example.com/coffee-news.jpg",
"link": "https://example.com/coffee-prices",
"position": 1,
"snippet": "Arabica futures climbed again this week.",
"source": "Example News",
"title": "Coffee prices hit a record high"
}
],
"credits_used": 1
}Search Google Scholar
Search academic literature on Google Scholar and get each paper's publication details, citation count, and PDF link where one is available.
Inputs
- Search query
querytextRequired - The topic, paper title, or author to search for.
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
Outputs
article_countnumber- How many papers were returned.
articleslist- The papers, each with its title, link, snippet, publication details, year, citation count, and PDF link.
credits_usednumber- How many Serper credits this search spent.
Example output
{
"article_count": 1,
"articles": [
{
"cited_by": 412,
"id": "paper-123",
"link": "https://example.com/paper",
"pdf_url": "https://example.com/paper.pdf",
"position": 1,
"publication_info": "A Researcher, B Researcher - Journal of Examples, 2019",
"snippet": "We review 42 randomized trials of caffeine.",
"title": "The effects of caffeine on alertness",
"year": 2019
}
],
"credits_used": 1
}Search Google Shopping
Search Google Shopping and get the matching products with their price, seller, delivery terms, and rating.
Inputs
- Search query
querytextRequired - The product to search for.
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
- Location
locationtext - Search from a specific place, which changes the local results Google shows. Write it the way Google does, most specific part first.
Outputs
credits_usednumber- How many Serper credits this search spent.
product_countnumber- How many products were returned.
productslist- The products, each with its title, link, seller, price, delivery, image, and rating.
Example output
{
"credits_used": 1,
"product_count": 1,
"products": [
{
"delivery": "Free delivery",
"image_url": "https://example.com/product.jpg",
"link": "https://example.com/product",
"offers": "12 offers",
"position": 1,
"price": "$13.86",
"product_id": "1234567890",
"rating": 4.6,
"rating_count": 16000,
"source": "Example Store",
"title": "Folgers Classic Roast Ground Coffee"
}
]
}Search Google Videos
Search Google Videos and get the matching videos with their title, channel, length, and link.
Inputs
- Search query
querytextRequired - What kind of video to search for.
- Number of results
num_resultsnumber - How many results to ask for, from 1 to 100. Defaults to 10. Google returns results in pages of ten, so the count you get back can be a little under what you ask for.From 1 to 100. Default:
10. - Page
pagenumber - Which page of results to return. 1 (the default) is the first page.At least 1. Default:
1. - Country
countrytext - Search as if from this country, as a two-letter country code. Defaults to the United States.
- Language
languagetext - Return results in this language, as a two-letter language code. Defaults to English.
- Published within
date_rangetext - Only return results published recently. Leave blank for any time.One of:
qdr:h(Past hour),qdr:d(Past 24 hours),qdr:w(Past week),qdr:m(Past month),qdr:y(Past year).
Outputs
credits_usednumber- How many Serper credits this search spent.
video_countnumber- How many videos were returned.
videoslist- The videos, each with its title, link, snippet, thumbnail, duration, source, channel, and date.
Example output
{
"credits_used": 1,
"video_count": 1,
"videos": [
{
"channel": "Example Channel",
"date": "1 week ago",
"duration": "8:12",
"image_url": "https://example.com/video-thumb.jpg",
"link": "https://example.com/watch",
"position": 1,
"snippet": "A short guide to grind size and water temperature.",
"source": "Example Video",
"title": "How to brew better coffee"
}
]
}