Content
Moderation API
Checks video, text, image, audio for any unwanted content
Our solutions can flexibly adapt
to your business’s values
Images
Check the quality and appropriateness of images to make your service even more attractive to users.
Show more
Text
Detect unpleasant or rude communication or mentions of prohibited products and services.
Show more
Video
Check videos for prohibited scenes and content that might interfere with a pleasant viewing.
Show more
Solution can detect the following classes
Spam
Content that isn't useful and aims to lure the reader to some internet resource.
Clickbait
A message that excites and arouses curiosity but doesn't give complete information.
Obscene
Rude expressions, including those censored with asterisks or dots.
Profanity
Insults addressed at any people or material objects.
Offensive content
Insults addressed at any people or material objects.
Erotica
Descriptions of sex or someone's appearance that are informational
in nature and don't mention gross physiological details.
in nature and don't mention gross physiological details.
Sex toys
Any mention of sex toys and the experience of using them, and reviews without physiological details.
Violence
Shocking or disgusting text. Calls for violence and cruelty. Encouraging hatred on grounds of gender, race, nationality, or association with any social group
Drugs
The mention of drugs in any context, including their promotion, descriptions of manufacture methods, and the use and sale of prohibited narcotic substances.
Quickstart: Use the Content
Moderator client API
Request
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
import requests
token = 'abacaba' # set real token here
api_url = 'https://api.membrace.ai/v1/item/process'
headers = {'Authorization': token}
# Text request
text = 'Hello world'
resp = requests.post(
api_url,
headers=headers,
json={
'request_id': '1',
'project_id': 'text',
'project_version': '1',
'content': {
'text': text
}
}
)
resp.raise_for_status()
print(resp.json())
Response
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
{
"request_id": "1",
"item_id": "safdhgjkl",
"project_id": <project_id>,
"project_version": "1",
"status": "SUCCESS",
"result": {
"alcohol": false,
"clean_text": true,
"clickbait": false,
"drugs": false,
"erotica": false,
"explicit": false,
"insult": false,
"insult_light": false,
"insult_strong": false,
"obscene_light": false,
"obscene_strong": false,
"personal_data": false,
"swearing": false
}
}
Request
1
2
3
4
5
6
7
8
9
10
11
12
curl -X POST \
-H 'Authorization: <token>' \
-H 'Content-Type: application/JSON' \
-d '{
"request_id": "1",
"project_id": <project_id>,
"project_version": "1",
"content": {
"text": "Example message"
}
}' \
https://api.membrace.ai/v1/item/process
Response
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
{
"request_id": "1",
"item_id": "safdhgjkl",
"project_id": <project_id>,
"project_version": "1",
"status": "SUCCESS",
"result": {
"alcohol": false,
"clean_text": true,
"clickbait": false,
"drugs": false,
"erotica": false,
"explicit": false,
"insult": false,
"insult_light": false,
"insult_strong": false,
"obscene_light": false,
"obscene_strong": false,
"personal_data": false,
"swearing": false
}
}
Have questions? Talk to an expert
Why working with Membrace?
Large and responsive crowd
We have over 2M workers distributed over the globe, so manual checks are available 24/7
Scalability
We can process any volume of data, including spikes caused by spam attacks, breaking news, flash mobs, or organic growth
Flexibility
Customizable moderation classes for brand-appropriate responses
Continuous fine-tuning
Our crowd force can handle any granularity of detail, and our models continue to learn from incoming user data
High accuracy
in unique cases
We apply a multitude of QA methods to both our human-in-the-loop moderators and our ML models
Real-time moderation
For rapidly growing, large-scale, and fast-paced businesses
