Contextual ad

segments & filters


Probability of
Probability of

Death or Injuries

Crime or Violence

67% 75%

Impression-level intelligence

Our API offers precise impression-level filtering of news content. Annotate and filter news content to match a wide array of brand standards.

Simple integration

Select models & filters

Pick the models that you want to filter for, and the domain categories you feel comfortable with. Stack socially equitable models on top to ensure you're not missing socially equitable content.

Send the URL or text

Process text content, or have us fetch the text content using our content scrapers especially designed for news websites.

Get predictions

Get probabilities for each news filter you selected.

Classify impressions

Follow our guidelines to decide if the impression is acceptable. Use our probability guidelines to optimize for either recall (max safety) or precision (max impressions).

Available Sensitive Filters

We’re adding new news filters every other week or so, but right now, our API classifies the following news content:

crime 1

Crime & violence

Everything from murder and battery to embezzling or tax fraud. If it’s illegal, or depicts violence or harm towards others, it is captured in this category.

military 2


The actions of state or nation owned military. Including torture, and policy-related decisions associated with military actions such as spending.

death 3

Death & injury

Death or injury not attributed to malice, such as an individual dying from cancer, or an athlete being injured. Includes obituaries.

sex 4


Content that is sexual in nature or includes sensitive sexual identifiers.

profanity 5


Language that your parents would tell you not to use. Includes thousands of clever variants such as $h!t.

drugs 6

Drugs & alcohol

Mentions of illegal drugs, alcohol and tobacco. Includes all the slang we could find from the dark edges of the Internet.

politics 7


An exhaustive classifier of all things political including campaigns, congress, and political actors.

antivax 8


Articles that use misinformation or disinformation to disparage the efficacy or use of vaccines.

Socially Equitable Inclusion Segments

Racial diversity

Detect positive, socially equitable news content. Our diversity filter identifies articles that talk about racial diversity and racial equality positively. For instance, an article may talk about the positive results of racial equality, or talk about the change needed to improve the quality of society. Current pro social segments include racial diversity, gender equality, and LGBT rights.

Use our socially conscious filters to allow advertisers to invest in diversity, equity and inclusion.

Domain-level filtering

We also offer exhaustive news domain exclusion lists to allow advertisers to avoid sensitive domains. Through a decade of academic journalism and media research we’ve studied nearly 3,000 news media websites and classified them for their:

  • Media type: newspapers, TV & radio stations, blogs and online-only media
  • Credibility: filter domains that have received bad fact-checking scores
  • Partisanship: filter hard right or hard left media sources

Want to give it a try?