POST
/
api
/
search_similar

Authorizations

x-api-token
string
headerrequired

API Key to authenticate requests.

To access the API, include your API key in the x-api-token header. To obtain your API key, complete the form or contact us directly.

Body

application/json
q
string
required

The keyword(s) to search for in articles. Query syntax supports logical operators (AND, OR, NOT) and wildcards:

  • For an exact match, use double quotes. For example, "technology news".
  • Use * to search for any keyword.
  • Use + to include and - to exclude specific words or phrases. For example, +Apple, -Google.
  • Use AND, OR, and NOT to refine search results. For example, technology AND (Apple OR Microsoft) NOT Google.

For more details, see Advanced querying.

search_in
string
default: title,content

The article fields to search in. To search in multiple fields, use a comma-separated string.

Example: "title, summary"

Note: The summary option is available if NLP is enabled in your plan.

Available options: title, summary, content.

include_similar_documents
boolean
default: false

If true, includes similar documents in the response.

similar_documents_number
integer
default: 5

The number of similar documents to return.

similar_documents_fields
string
default: title,content

The fields to consider for finding similar documents.

predefined_sources

Predefined top sources per country.

Format: start with the word top, followed by the number of desired sources, and then the two-letter country code ISO 3166-1 alpha-2.

Examples:

  • "top 100 US"
  • "top 33 AT"
  • "top 5 GB"

Multiple countries can be specified with custom numbers as a comma-separated string or an array of strings.

Examples:

  • "top 50 US, top 20 GB"
  • ["top 50 US", "top 20 GB"]
sources

One or more news sources to narrow down the search. The format must be a domain URL. Subdomains, such as finance.yahoo.com, are also acceptable. To specify multiple sources, use a comma-separated string or an array of strings.

Examples:

  • "nytimes.com, theguardian.com"
  • ["nytimes.com", "theguardian.com"]
not_sources

The news sources to exclude from the search. To exclude multiple sources, use a comma-separated string or an array of strings.

Examples:

  • "cnn.com, wsj.com"
  • ["cnn.com", "wsj.com"]
lang

The language(s) of the search. The only accepted format is the two-letter ISO 639-1 code. To select multiple languages, use a comma-separated string or an array of strings.

Examples:

  • "en,es"
  • ["en", "es"]

To learn more, see Enumerated parameters > Language.

not_lang

The language(s) to exclude from the search. The accepted format is the two-letter ISO 639-1 code. To exclude multiple languages, use a comma-separated string or an array of strings.

Examples:

  • "fr,de"
  • ["fr", "de"]

To learn more, see Enumerated parameters > Language.

countries

The countries where the news publisher is located. The accepted format is the two-letter ISO 3166-1 alpha-2 code. To select multiple countries, use a comma-separated string or an array of strings.

Examples:

  • "US,CA"
  • ["US", "CA"]

To learn more, see Enumerated parameters > Country.

not_countries

The publisher location countries to exclude from the search. The accepted format is the two-letter ISO 3166-1 alpha-2 code. To exclude multiple countries, use a comma-separated string or an array of strings.

Examples:

  • "UK,FR"
  • ["UK", "FR"]

To learn more, see Enumerated parameters > Country.

from_

The starting point in time to search from. Accepts date-time strings in ISO 8601 format and plain text strings. The default time zone is UTC.

Formats with examples:

  • YYYY-mm-ddTHH:MM:SS: 2024-07-01T00:00:00
  • YYYY-MM-dd: 2024-07-01
  • YYYY/mm/dd HH:MM:SS: 2024/07/01 00:00:00
  • YYYY/mm/dd: 2024/07/01
  • English phrases: 1 day ago, today

Note: By default, applied to the publication date of the article. To use the article's parse date instead, set the by_parse_date parameter to true.

to_

The ending point in time to search up to. Accepts date-time strings in ISO 8601 format and plain text strings. The default time zone is UTC.

Formats with examples:

  • YYYY-mm-ddTHH:MM:SS: 2024-07-01T00:00:00
  • YYYY-MM-dd: 2024-07-01
  • YYYY/mm/dd HH:MM:SS: 2024/07/01 00:00:00
  • YYYY/mm/dd: 2024/07/01
  • English phrases: 1 day ago, today

Note: By default, applied to the publication date of the article. To use the article's parse date instead, set the by_parse_date parameter to true.

by_parse_date
boolean
default: false

If true, the from_ and to_ parameters use article parse dates instead of published dates. Additionally, the parse_date variable is added to the output list for each article.

published_date_precision
enum<string>

The precision of the published date. There are three types:

  • full: The day and time of an article is correctly identified with the appropriate timezone.
  • timezone unknown: The day and time of an article is correctly identified without timezone.
  • date: Only the day is identified without an exact time.
Available options:
full,
timezone unknown,
date
sort_by
enum<string>
default: relevancy

The sorting order of the results. Possible values are:

  • relevancy: The most relevant results first.
  • date: The most recently published results first.
  • rank: The results from the highest-ranked sources first.
Available options:
relevancy,
date,
rank
ranked_only
boolean
default: true

If true, limits the search to sources ranked in the top 1 million online websites. If false, includes unranked sources which are assigned a rank of 999999.

from_rank
integer
default: 1

The lowest boundary of the rank of a news website to filter by. Range: 1 to 999999, where a lower rank indicates a more popular source. If you set this to 100, the API includes sources ranked 100 or higher.

to_rank
integer
default: 999999

The highest boundary of the rank of a news website to filter by. Range: 1 to 999999, where a lower rank indicates a more popular source. If you set this to 100, the API includes sources ranked 100 or lower.

is_headline
boolean

If true, only returns articles that were posted on the home page of a given news domain.

is_opinion
boolean

If true, returns only opinion pieces. If false, excludes opinion-based articles and returns news only.

is_paid_content
boolean

If false, returns only articles that have publicly available complete content. Some publishers partially block content, so this setting ensures that only full articles are retrieved.

parent_url

The categorical URL(s) to filter your search. To filter your search by multiple categorical URLs, use a comma-separated string or an array of strings.

Examples:

  • "wsj.com/politics,wsj.com/tech"
  • ["wsj.com/politics", "wsj.com/tech"]
all_links

The complete URL(s) mentioned in the article. For multiple URLs, use a comma-separated string or an array of strings.

Examples:

  • "https://aiindex.stanford.edu/report/, https://www.stateof.ai/"
  • ["https://aiindex.stanford.edu/report/", "https://www.stateof.ai/"]

For more details, see Search by URL.

all_domain_links

The domain(s) mentioned in the article. For multiple domains, use a comma-separated string or an array of strings.

Examples:

  • "who.int, nih.gov"
  • ["who.int", "nih.gov"]

For more details, see Search by URL.

word_count_min
integer

The minimum number of words an article must contain. To be used for avoiding articles with small content.

word_count_max
integer

The maximum number of words an article can contain. To be used for avoiding articles with large content.

page
integer
default: 1

The page number to scroll through the results. This parameter is used to paginate: scroll through results because one API response cannot return more than 1000 articles.

page_size
integer
default: 100

The number of articles to return per page. Range: 1 to 1000.

include_nlp_data
boolean
default: false

If true, includes an NLP layer with each article in the response. This layer provides enhanced information such as theme classification, article summary, sentiment analysis, tags, and named entity recognition.

The NLP layer includes:

  • Theme: General topic of the article.
  • Summary: A concise overview of the article content.
  • Sentiment: Separate scores for title and content (range: -1 to 1).
  • Named entities: Identified persons (PER), organizations (ORG), locations (LOC), and miscellaneous entities (MISC).
  • IPTC tags: Standardized news category tags.
  • IAB tags: Content categories for digital advertising.

Note: The include_nlp_data parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

has_nlp
boolean
default: false

If true, filters the results to include only articles with an NLP layer. This allows you to focus on articles that have been processed with advanced NLP techniques.

Note: The has_nlp parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

theme

Filters articles based on their general topic, as determined by NLP analysis. To select multiple themes, use a comma-separated string or an array of strings.

Examples:

  • "Finance, Tech"
  • ["Finance", "Tech"]

Note: The theme parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

Available options: Business, Economics, Entertainment, Finance, Health, Politics, Science, Sports, Tech, Crime, Financial Crime, Lifestyle, Automotive, Travel, Weather, General.

not_theme

Inverse of the theme parameter. Excludes articles based on their general topic, as determined by NLP analysis. To exclude multiple themes, use a comma-separated string or an array of strings.

Examples:

  • "Crime, Tech"
  • ["Crime", "Tech"]

Note: The not_theme parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

ner_name
string

The name of person, organization, location, product or other named entity to search for. To specify multiple names use a comma-separated string.

Example: "Tesla, Amazon"

title_sentiment_min
number

Filters articles based on the minimum sentiment score of their titles.

Range is -1.0 to 1.0, where:

  • Negative values indicate negative sentiment.
  • Positive values indicate positive sentiment.
  • Values close to 0 indicate neutral sentiment.

Note: The title_sentiment_min parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

title_sentiment_max
number

Filters articles based on the maximum sentiment score of their titles.

Range is -1.0 to 1.0, where:

  • Negative values indicate negative sentiment.
  • Positive values indicate positive sentiment.
  • Values close to 0 indicate neutral sentiment.

Note: The title_sentiment_max parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

content_sentiment_min
number

Filters articles based on the minimum sentiment score of their content.

Range is -1.0 to 1.0, where:

  • Negative values indicate negative sentiment.
  • Positive values indicate positive sentiment.
  • Values close to 0 indicate neutral sentiment.

Note: The content_sentiment_min parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

content_sentiment_max
number

Filters articles based on the maximum sentiment score of their content.

Range is -1.0 to 1.0, where:

  • Negative values indicate negative sentiment.
  • Positive values indicate positive sentiment.
  • Values close to 0 indicate neutral sentiment.

Note: The content_sentiment_max parameter is only available if NLP is included in your subscription plan.

To learn more, see NLP features.

iptc_tags

Filters articles based on IPTC (International Press Telecommunications Council) media topic tags. To specify multiple IPTC tags, use a comma-separated string or an array of strings.

Examples:

  • "20000199, 20000209"
  • ["20000199", "20000209"]

Note: The iptc_tags parameter is only available if tags are included in your subscription plan.

To learn more, see IPTC Media Topic NewsCodes.

not_iptc_tags

Inverse of the iptc_tags parameter. Excludes articles based on IPTC (International Press Telecommunications Council) media topic tags. To specify multiple IPTC tags to exclude, use a comma-separated string or an array of strings.

Examples:

  • "20000205, 20000209"
  • ["20000205", "20000209"]

Note: The not_iptc_tags parameter is only available if tags are included in your subscription plan.

To learn more, see IPTC Media Topic NewsCodes.

Response

200 - application/json

The response model for a successful search for similar articles.

status
string
default: ok

The status of the response.

total_hits
integer
required

The total number of articles matching the search criteria.

page
integer
required

The current page number of the results.

total_pages
integer
required

The total number of pages available for the given search criteria.

page_size
integer
required

The number of articles per page.

articles
object[]
required

A list of articles matching the search criteria.

user_input
object
required

The user input parameters for the search.

Was this page helpful?

Search similar articlesRetrieve sources