For the complete documentation index, see llms.txt. This page is also available as Markdown.

Search Engine Scrapers

Reference for Web API's Google and Bing scraper endpoints: parameters and delivery modes for every endpoint.

Scraper
Endpoints

/google/search, /google/ads, /google/aimode, /google/shopping/search, /google/shopping/product, /google/maps, /google/travel/hotels, /google/lens, /google/trends/explore, /google

/bing/search, /bing

Google

POST /v1/scrape/google/search

Returns organic results, ads, and the SERP features that appeared for a query.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Search phrase, as a user would type it

aomd

string

no

Undocumented by Google. Forwarded as sent

cr

string

no

Country restriction. Forwarded verbatim

disable_scripts

boolean

no

Blocks page scripts

domain

string

no

Google TLD to search: com, co.uk, de. Selects the national index

expand_aio

boolean

no

Expands the AI Overview block instead of returning it collapsed

filter

integer

no

Google's duplicate-content filter. 0 turns it off and returns near-duplicate results

fpstate

string

no

Page state. Forwarded verbatim

limit_per_page

array of objects

no

Per-page result counts, e.g. [{"page": 1, "limit": 20}]. Use instead of pages when only page 1 matters

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

nfpr

string

no

Suppresses Google's spelling auto-correction

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

results_language

string

no

Restricts results to one language, e.g. lt. Filters, does not translate

safe_search

boolean

no

Google's SafeSearch filter. A boolean here, a string on Bing

start_page

integer

no

First page to fetch. Default 1

tbm

string

no

Search vertical: images, shopping, video, books. Forwarded verbatim

tbs

string

no

Time and quality filters, e.g. past 24 hours. Forwarded verbatim

udm

string

no

Google's UI-mode selector. Forwarded verbatim

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/ads

Returns the paid block only: advertisers, order, and ad copy.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Search phrase the ads are fetched for

adstest

boolean

no

Returns ads from Google's test ad network instead of live inventory

disable_scripts

boolean

no

Blocks page scripts

domain

string

no

Google TLD to search: com, co.uk, de. Selects the national index

expand_aio

boolean

no

Expands the AI Overview block instead of returning it collapsed

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

start_page

integer

no

First page to fetch. Default 1

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/aimode

Returns Google's AI Mode answer, separate from the AI Overview returned by /google/search.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

The question to ask Google AI Mode

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

yes

Required: true

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

Responses may differ between identical requests.

POST /v1/scrape/google/shopping/search

Returns Google Shopping listings with prices.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Search phrase

domain

string

no

Google TLD to search: com, co.uk, de. Selects the national index

limit

integer

no

Number of results to return

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

max_price

integer

no

Highest price to include, in the storefront's currency

min_price

integer

no

Lowest price to include, in the storefront's currency

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

sort_by

string

no

r relevance (default), rv rating high to low, p price low to high, pd price high to low

start_page

integer

no

First page to fetch. Default 1

tbs

string

no

Time and quality filters, e.g. past 24 hours. Forwarded verbatim

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/shopping/product

Returns one Google Shopping product.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Shopping product ID

domain

string

no

Google TLD to search: com, co.uk, de. Selects the national index

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/maps

Returns Google Maps local listings.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Place or category search phrase, e.g. dentist vilnius

domain

string

no

Google TLD to search: com, co.uk, de. Selects the national index

hotel_dates

string

no

Stay dates. Without it, Google's default stay window is used

hotel_occupancy

integer

no

Guests in the room

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

start_page

integer

no

First page to fetch. Default 1

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Required with output: ["json"] — this endpoint has no built-in parser. {"prompt": string} or {"schema": object}, never both

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/travel/hotels

Returns hotel listings priced for the given dates and occupancy.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Hotel name or destination

adults

integer

no

Adult guests

children

integer

no

Child guests

domain

string

no

Google TLD to search: com, co.uk, de. Selects the national index

hotel_classes

array of integers

no

Star ratings to include, e.g. [4, 5]

hotel_dates

string

no

Stay dates. Without it, Google's default stay window is used

hotel_occupancy

integer

no

Guests in the room

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

start_page

integer

no

First page to fetch. Default 1

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/lens

Returns visual-match results for an image URL.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Image URL to match. Not search text

keywords

string

no

Text hint to narrow the visual matches

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/google/trends/explore

Returns relative search interest over time for a term.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Term to chart

category_id

string

no

Google Trends category ID

date_from

string

no

Start of the window, e.g. 2026-01-01

date_to

string

no

End of the window, e.g. 2026-06-30

search_type

string

no

Trends vertical, e.g. web, image, news, YouTube, or shopping

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Required with output: ["json"] — this endpoint has no built-in parser. {"prompt": string} or {"schema": object}, never both

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

Interest values are relative, normalized within the requested window: widening date_from rescales every point.

POST /v1/scrape/google

Returns parsed fields for a Google page from a URL you provide.

This endpoint can also collect and parse Travel: Hotels details. See /google/travel/hotels.

Modes: sync, async, batch

Name
Type
Required
Description

url

string

yes

A Google URL

adstest

boolean

no

Returns ads from Google's test ad network instead of live inventory

adults

integer

no

Adult guests

aomd

string

no

Undocumented by Google. Forwarded as sent

category_id

string

no

Google Trends category ID

children

integer

no

Child guests

cr

string

no

Country restriction. Forwarded verbatim

date_from

string

no

Start of the window, e.g. 2026-01-01

date_to

string

no

End of the window, e.g. 2026-06-30

disable_scripts

boolean

no

Blocks page scripts

expand_aio

boolean

no

Expands the AI Overview block instead of returning it collapsed

filter

integer

no

Google's duplicate-content filter. 0 turns it off and returns near-duplicate results

fpstate

string

no

Page state. Forwarded verbatim

hotel_classes

array of integers

no

Star ratings to include, e.g. [4, 5]

hotel_dates

string

no

Stay dates. Without it, Google's default stay window is used

hotel_occupancy

integer

no

Guests in the room

limit

integer

no

Number of results to return

limit_per_page

array of objects

no

Per-page result counts, e.g. [{"page": 1, "limit": 20}]. Use instead of pages when only page 1 matters

locale

string

no

Interface language, e.g. en_US. Affects wording of SERP features, not ranking

max_price

integer

no

Highest price to include, in the storefront's currency

min_price

integer

no

Lowest price to include, in the storefront's currency

nfpr

string

no

Suppresses Google's spelling auto-correction

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

results_language

string

no

Restricts results to one language, e.g. lt. Filters, does not translate

safe_search

boolean

no

Google's SafeSearch filter. A boolean here, a string on Bing

search_type

string

no

Trends vertical, e.g. web, image, news, YouTube, or shopping

sort_by

string

no

r relevance (default), rv rating high to low, p price low to high, pd price high to low

start_page

integer

no

First page to fetch. Default 1

tbm

string

no

Search vertical: images, shopping, video, books. Forwarded verbatim

tbs

string

no

Time and quality filters, e.g. past 24 hours. Forwarded verbatim

udm

string

no

Google's UI-mode selector. Forwarded verbatim

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code of the request origin. Independent of domain

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

Set a parameter in the body or in the URL's query string, not both.

Bing

POST /v1/scrape/bing/search

Returns Bing organic results for a query.

Modes: sync, async, batch

Name
Type
Required
Description

query

string

yes

Search phrase

domain

string

no

Bing TLD

locale

string

no

Interface language, e.g. en_US

pages

integer

no

Consecutive pages to fetch from start_page. Each page is a separate fetch and charge

safe_search

string

no

Safe-search level. A string here, a boolean on Google

start_page

integer

no

First page to fetch. Default 1

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code to fetch from, e.g. DE

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

POST /v1/scrape/bing

Returns parsed fields for a Bing page from a URL you provide.

Modes: sync, async, batch

Name
Type
Required
Description

url

string

yes

A Bing URL

locale

string

no

Interface language, e.g. en_US

output

array of strings

no

Formats to return: markdown, html, json, screenshot. Default ["html"]. json returns parsed fields. screenshot requires run_js: true

json

object

no

Custom extraction: {"prompt": string} or {"schema": object}, never both. Overrides the built-in parser

location

string

no

Two-letter country code to fetch from, e.g. DE

device

string

no

desktop (default) or mobile

run_js

boolean

no

Executes page JavaScript before capture. Slower — use a client timeout of 120 seconds or more

callback_url

string

no

Webhook URL for the finished job. Async only. Malformed URLs are rejected with 400

storage

object

no

Delivers the result to a bucket instead of the response: {"type": string, "url": string}. type is s3, s3_gzip, s3_compatible, gcs, or tos. Async only. url is not validated

Last updated

Was this helpful?