Premium Feature Endpoint
Multi Claim Detection & Verification
This endpoint enables you to detect and verify content claims.
/custom/api/v2/detect/verify/claims
If no response is received here due to the extended generation time required by this API, we suggest utilizing a local API platform such as Postman or directly accessing the API through your project.
Headers
Input the Authorization Bearer Token to authorize your API request.
Body
This input field requires the user to input the blog for detecting the claims.
This input field requires the user to input the blog for detecting the claims.
This input field requires the user to input headline of blog, or blog topic, the more accurate the topic, the better the algorithm in fact checking
Response
Indicates whether the call was successful. It prompts success if working, otherwise not.
This is an array which contains objects having the following keys:
- 1. claim: displays the claim text that needs to be fact checked.
- 2. factuality: Indicates whether the claim is a factually true or not.
- 3. error: what is not true about the claim.
- 4. correction: correction that needs to be made to the claim.
- 5. reasoning: reasoning behind the assessment.
- 6. relevant_links: a list of relevant sources containing fact, link and score
curl --location --request POST 'https://api-az.longshot.ai/custom/api/v2/detect/verify/claims'
--header 'Content-Type: application/json'
--header 'Authorization: Bearer <token>'
--data-raw '{
"blog": "<Blog you want to detect claims from>",
"url": "<URL you want to detect claims from>",
"blog_headline_topic: "<headline of Blog, or Blog Topic, the more accurate the topic, the better the algorithm in fact checking>"
}'
{
"status": "Success",
"fact_check_results": [
{
"claim": "The motivations behind people's choice to shop online are diverse, ranging from the convenience it offers to the competitive prices available.",
"reasoning": "The evidence provided does not contain any readable or understandable text that can be used to verify the statement. The text in the evidence appears to be corrupted or encoded, making it impossible to extract any information from it.",
"factuality": "Need more information",
"relevant_links": [
{
"fact": "<Fact cited in fact check>",
"link": "<url to source of fact>",
"score": 0.1234
},
{
"fact": "<Fact cited in fact check>",
"link": "<url to source of fact>",
"score": 0.1234
}
],
"error": "The evidence provided is unreadable and cannot be used to verify the statement.",
"correction": "No correction can be provided as the evidence is unreadable."
}
]
}