Get paginated sentiment sources
Returns the sources cited by sentiment answers, grouped by domain or listed per URL, with the sentiment split each source contributed to. Supports server-side filtering and sorting, and returns the filter option values needed to build multi-select dropdowns. Responses are cached for 60 seconds.
API Key for programmatic access
In: header
Path Parameters
UUID of the domain
uuidQuery Parameters
Page number (1-based)
1 <= value1Items per page (1-1000, default 50)
1 <= value <= 100050Start of the report date range (ISO 8601)
End of the report date range (ISO 8601)
Global filter: LLM models the report was run with (repeat the param, or send one comma-separated value)
Group rows by source domain, or list individual cited URLs (default: domain)
Value in
- "domain"
- "url"
Multi-select filter on LLM model (repeat the param, or send one comma-separated value)
Filter by exact prompt texts (use repeated params, not comma-separated)
Partial match on the source/domain column
Sort direction (default: desc)
Value in
- "asc"
- "desc"
Restrict to one market/language pair
Multi-select filter on sentiment (comma-separated)
Field to sort by (default: count)
Value in
- "source"
- "count"
- "sentiment"
Response Body
application/json
curl -X GET "https://example.com/api/domains/3fa85f64-5717-4562-b3fc-2c963f66afa6/sources/sentiment"{ "data": [ { "id": "3fa85f64-5717-4562-b3fc-2c963f66afa6", "domain": "g2.com", "url": "https://g2.com/products/nike/reviews", "title": "Nike Reviews 2026", "prompts": [ "How do customers feel about Nike?" ], "models": [ "chatgpt", "claude" ], "count": 12, "sentiments": [ "positive", "neutral" ], "dominantSentiment": "positive" } ], "total": 142, "page": 1, "limit": 50, "totalPages": 3, "hasNext": true, "hasPrev": false, "filterOptions": { "models": [ { "value": "chatgpt", "label": "GPT-5", "score": 72 } ], "prompts": [ { "value": "chatgpt", "label": "GPT-5", "score": 72 } ], "sentiments": [ "positive", "neutral", "negative" ] }}