Detect Weapons in Images and Videos at scale

Detect firearms, knives, blades and more. Build smart filters based on the weapon's context and threat level with our automated APIs.

Have questions? Talk to sales

Powering some of the most innovative platforms worldwide

Detect Firearms and determine the level of threat

Detect guns and rifles with the Weapon API. Apply smart rules adapted to the context of the firearm and in line with your community guidelines. You can block any firearm, or implement adaptive filters depending on how the weapon is used. Here are the available categories:


Level 0

Immediate threat

Level 1

Rifle immediate threat
Aiming, no threat

Level 2

aiming at target with rifle
In hand not aiming

Level 3

rifle in hand, not aiming
Not in hand

Level 4

gun not in hand

Adapt your filters to the nature of the image

Use our proprietary Image Type models to determine if a given image or scene is natural or animated:

  • Real-life: situations with actual weapons.
  • In-game: scenes extracted from video games and that are perceived by viewers as such.
  • Illustrations: weapons featured in drawings, paintings, comics or sketches.
  • Computer-generated: other types of computer-generated weapons that look distinct from the "real thing".

Differentiate firearms based on their nature

Real rifle

Real-life scene

Real rifle
Animated rifle

In-game scene

animated gun

Detect different types of Knives along with their context

Detect utility knives, combat knives, cleavers, and daggers with the Weapon API. Adapt your filtering to your users and your community guidelines. You can block any knife, or implement adaptive filters depending on the type of knife and how it is being used. Here are a few categories:

Self-harm with knife
Rifle immediate threat
Gore & Knife
Rifle immediate threat
Non culinary knife in use
Non culinary knife in use
Non culinary knife, not in use
Non culinary knife not in use
Culinary knife in use
Culinary knife in use
Culinary knife, not in use
Culinary knife not in use

Combine with the Violence, Gore, and Self-Harm models, to name just a few

Sightengine provides you with 40+ content moderation categories that you can pick and choose from to detect any type of unwanted content.
Here are a few examples:

Violence Detection

Physical violence, battery, assault as well hanging, executions, strangling, restraining and more...

Gore and Horrific Content Detection

Displays including blood, open wounds, mutilation, death and other graphic violence or horrific imagery.

Self-Harm Detection

Self-injury, threats targeted to oneself, self-directed cuts and other displays of self-harm or intention to commit suicide.

See other models

Quick and Easy to Integrate

Our API was designed by developers for developers. We have worked hard to make it simple and quick to integrate with. Access our detailed documentation, quickstart guides, easy-to-integrate SDKs or talk to our live support whenever needed.
# copy and paste this in your terminal
curl -X GET -G '' \
     -d 'url=' \
     -d 'models=wad'
# install the SDK with "pip install sightengine"
client = SightengineClient('{api_user}', '{api_secret}')
output = client.check('weapon') \
// install the SDK with "composer require sightengine/client-php"
use \Sightengine\SightengineClient;
$client = new SightengineClient('{api_user}', '{api_secret}');
$check = $client->check(['weapon']);
$output = $check->set_url('');
// install the SDK with "npm install sightengine --save"
var sightengine = require('sightengine')('{api_user}', '{api_secret}');
sightengine.check(['weapon']).set_url('').then(function(result) {
  // The API response (result)
}).catch(function(err) {
  // Handle error

See why the world's best companies are using Sightengine

Empower your business with Powerful Content Analysis and Filtering.