Sentiment

POST /v2/reports/sentiment

cURL

curl -X POST https://api.tryprofound.com/v2/reports/sentiment \
  -H "X-API-Key: <your-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "category_id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "asset": "Profound",
    "start_date": "2026-06-09",
    "end_date": "2026-06-15"
  }'

200

{
  "info": {
    "total_results": 1,
    "count": 1,
    "next_cursor": null,
    "asset": "Profound",
    "models": ["ChatGPT", "Google Gemini", "Perplexity", "Claude", "..."],
    "metrics": ["positive_sentiment", "negative_sentiment"],
    "start_date": "2026-06-09",
    "end_date": "2026-06-15",
    "filter": null
  },
  "data": [
    {
      "positive_sentiment": 79.1,
      "negative_sentiment": 20.9
    }
  ]
}

Per-brand sentiment as percentages from either AI responses or the pages those responses cite. asset is required because sentiment is calculated for one brand at a time.

• source: "response" (default): sentiment in AI responses. Metrics are positive_sentiment, negative_sentiment, and opt-in occurrence.
• source: "citation": one row per cited page, with positive_sentiment, negative_sentiment, and model-balanced citation_share. It defaults to sorting by citation_share; explicitly set sort.field to positive_sentiment to sort by page sentiment instead.
• Comparison dates: add comparison_start_date and comparison_end_date. Response rows include previous sentiment metrics. Citation rows include the page’s previous citation_share.

Response sentiment supports up to two group_by dimensions from topic, region, model, prompt, persona, tag, theme, claim, run, and competitor, plus date. Citation sentiment is page-only and rejects group_by; use filters to narrow the page population.

Both sources accept prompt-layer filters such as model, topic, region, persona, prompt, and tag, plus top-level theme / claim filters. Citation sentiment additionally accepts these top-level filters:

• citation_category: is or in. Values can be built-in categories (owned, competition, social, earned_media, earned_institutions, pr_wire, other) or a custom category value.
• page: contains_case_insensitive with one non-empty value. This searches the normalized page URL.

include_cited_websites: true applies only to response sentiment grouped by theme and/or claim.

Citation sentiment describes the cited page. A model filter changes which models’ citations contribute to page eligibility and citation share; it does not turn page sentiment into a model-specific sentiment score.

New to the v2 reports? See Filtering & concepts for the shared request shape, filter tree, grouping, and pagination.

Streaming (SSE) variant (same body, /stream)

POST /v2/reports/sentiment/stream supports both response and citation sentiment and returns Server-Sent Events: one summary event (the info block), then one result event per row. limit/cursor are ignored; it returns everything by default. Pass max_results to cap. With source: "citation", each result is the same page-level shape returned in data by the paginated endpoint.

Response (text/event-stream)

event: summary
data: { ...the info block... }

event: result    ← one per row, same shape as data[] above
data: {"theme": {"id": "a14e0c2d", "name": "Analytics"}, "rank": 1, "positive_sentiment": 95.4, "negative_sentiment": 4.6}

cURL

curl -X POST https://api.tryprofound.com/v2/reports/sentiment \
  -H "X-API-Key: <your-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "category_id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "asset": "Profound",
    "start_date": "2026-06-09",
    "end_date": "2026-06-15"
  }'

200

{
  "info": {
    "total_results": 1,
    "count": 1,
    "next_cursor": null,
    "asset": "Profound",
    "models": ["ChatGPT", "Google Gemini", "Perplexity", "Claude", "..."],
    "metrics": ["positive_sentiment", "negative_sentiment"],
    "start_date": "2026-06-09",
    "end_date": "2026-06-15",
    "filter": null
  },
  "data": [
    {
      "positive_sentiment": 79.1,
      "negative_sentiment": 20.9
    }
  ]
}

Citation-page example

This request returns owned pages containing tryprofound.com, ordered by their current citation share. Combine citation-layer leaves with other filters under a top-level and.

cURL

curl -X POST https://api.tryprofound.com/v2/reports/sentiment \
  -H "X-API-Key: <your-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "category_id": "182bd5e5-6e1a-4fe4-a799-aa6d9a6ab26e",
    "asset": "Profound",
    "start_date": "2026-06-09",
    "end_date": "2026-06-15",
    "source": "citation",
    "metrics": ["positive_sentiment", "negative_sentiment", "citation_share"],
    "sort": {"field": "citation_share", "dir": "desc"},
    "filter": {
      "and": [
        {"field": "citation_category", "op": "is", "value": "owned"},
        {"field": "page", "op": "contains_case_insensitive", "value": "tryprofound.com"}
      ]
    },
    "limit": 10
  }'

200

{
  "info": {
    "total_results": 1,
    "count": 1,
    "next_cursor": null,
    "asset": "Profound",
    "models": ["ChatGPT", "Google Gemini", "Perplexity", "..."],
    "metrics": ["positive_sentiment", "negative_sentiment", "citation_share"],
    "start_date": "2026-06-09",
    "end_date": "2026-06-15",
    "filter": {
      "and": [
        {"field": "citation_category", "op": "is", "value": "owned"},
        {"field": "page", "op": "contains_case_insensitive", "value": "tryprofound.com"}
      ]
    }
  },
  "data": [
    {
      "page": "tryprofound.com/",
      "rank": 1,
      "positive_sentiment": 100.0,
      "negative_sentiment": 0.0,
      "citation_share": 1.42
    }
  ]
}

Authorizations

APIKeyHeaderBearerAuth

X-API-Key · string · header · required

Body

application/json

category_id · string<uuid> · required

asset · string · required

The brand name to analyze (sentiment is extracted on name, not id).

start_date · string · required

YYYY-MM-DD, ET, inclusive

end_date · string · required

YYYY-MM-DD, ET, inclusive

comparison_start_date · string | null

YYYY-MM-DD, ET, inclusive (with end).

comparison_end_date · string | null

YYYY-MM-DD, ET, inclusive (with start).

source · enum<string> · default: response

Available options: response, citation

group_by · enum<string>[]

Available options: date, model, topic, region, prompt, persona, tag, theme, claim, run, competitor

metrics · enum<string>[] | null

Available options: positive_sentiment, negative_sentiment, occurrence, citation_share

interval · enum<string> · default: day

Available options: day, week, month

filter · FilterNode · object | null

A leaf (field/op/value) or an and/or/not group.

Show child attributes

sort · SortSpec · object

Show child attributes

include_cited_websites · boolean · default: false

Return cited websites per row (only when grouping by theme/claim).

limit · integer | null

Page size; default 10, max 50.

Required range: 0 < x <= 50

max_results · integer | null

Stream endpoint only: cap the number of streamed rows (default: all).

Required range: x > 0

cursor · string | null

Response

200

application/json

Successful Response

info · SentimentV2Info · object · required

Show child attributes

data · SentimentRow · object[] · required

Show child attributes