Skip to main content
POST
/
api
/
analytics
/
models
/
drill-down
/
sources
Get top sources for a specific AI model
curl --request POST \
  --url https://api.mentionlab.io/api/analytics/models/drill-down/sources \
  --header 'Content-Type: application/json' \
  --header 'x-project-id: <x-project-id>' \
  --data '
{
  "model": "gpt-4o",
  "startDate": "2025-01-01",
  "endDate": "2025-02-01",
  "country": "<string>",
  "language": "<string>",
  "models": [
    "<string>"
  ],
  "hasSources": true,
  "hasShopping": true,
  "queryTagIds": [
    "<string>"
  ],
  "execTagIds": [
    "<string>"
  ],
  "queryTagMode": "or",
  "execTagMode": "or",
  "groupByEntityGroup": false,
  "entityTypes": [
    "owned",
    "primary",
    "competitor"
  ],
  "entityIds": [
    "<string>"
  ]
}
'
{
  "totalResponses": 500,
  "sources": [
    {
      "domain": "example.com",
      "resultCount": 50,
      "citationRate": 10,
      "pageCount": 75
    }
  ]
}

Headers

​
x-project-id
string
required

Project ID to specify the project context

Body

application/json
​
model
string
required

The specific AI model to drill down into (e.g. "gpt-4o")

Example:

"gpt-4o"

​
startDate
string

Start date (inclusive)

Example:

"2025-01-01"

​
endDate
string

End date (exclusive)

Example:

"2025-02-01"

​
country
string

Filter by country code

​
language
string

Filter by language code

​
models
string[]

Filter by AI models

​
hasSources
boolean

Filter by responses that have sources

​
hasShopping
boolean

Filter by responses that have shopping products

​
queryTagIds
string[]

Filter by query tag IDs

​
execTagIds
string[]

Filter by execution tag IDs

​
queryTagMode
enum<string>
default:or

Query tag matching mode: "or" matches ANY tag (default), "and" matches ALL tags.

Available options:
and,
or
​
execTagMode
enum<string>
default:or

Execution tag matching mode: "or" matches ANY tag (default), "and" matches ALL tags.

Available options:
and,
or
​
groupByEntityGroup
boolean
default:false

When true, entities belonging to the same entity_group are collapsed into a single row. Uses COUNT(DISTINCT ai_response_id) so that two group members present in the same response count as one (visibility cannot exceed 100%).

​
entityTypes
enum<string>[]

Filter which entity types to include in results. Accepted values: "owned", "primary", "competitor". Defaults to all types when omitted.

Available options:
owned,
primary,
competitor
Example:
["owned", "primary", "competitor"]
​
entityIds
string[]

Override entity IDs (defaults to primary entities)

Response

200 - application/json
​
totalResponses
number
required

Total number of AI responses for the selected model in the filtered scope (denominator for citationRate).

Example:

500

​
sources
object[]
required

Top source domains for the selected model, ordered by resultCount descending.