Moderations

Identify Safety

FanarGuard gives each prompt–response pair safety and cultural-awareness scores, allowing moderation thresholds to be tailored to the deployment.<br/>For our definition of cultural awareness, refer to https://arxiv.org/abs/2511.18852.

post/v1/moderations

Request

  • Base URL: https://api.fanar.qa (the document declares no server; this is the origin it was published from)
  • URL: https://api.fanar.qa/v1/moderations
  • Auth: HTTP bearer

Request body

model'Fanar-Guard-2' required
promptstring required

The prompt.

responsestring required

The model's response to the prompt.

Example request

{
  "model": "Fanar-Guard-2",
  "prompt": "What is the weather?",
  "response": "The weather is sunny today."
}

Response

Successful Response

safetynumber required

The safety score for general safety aspects such as toxicity, violence, self-harm, etc.

cultural_awarenessnumber required

The cultural awareness score for aspects such as stereotypes, insensitive content, and bias.

Changes

No recorded changes to this operation across all 1 revision of this API.