Top-N leaderboard | GetMint API

GET/domains/{domainId}/topics/{topicId}/visibility/mentions

The full brand-mention leaderboard behind the top-mentions widget — a paginated endpoint returning every mention ranked by count, each with its percentage share of voice. The dashboard renders it as this donut (demo data):

Share of Voice

22.2%Nike

Adidas27.4%
Nike22.2%
Puma13.9%
New Balance10.3%
Asics8.1%
Others18.1%

How this example works

percentage is share of voice — the mention's share of all brand mentions in the window (the column sums toward 100%), not a per-model figure.
Pagination is 1-indexed. limit default: 20 max 100 (higher values are rejected with a 400, not clamped), page starts at 1; rank is (page - 1) * limit + index + 1.
Your own brand appears here too. For a guaranteed own-brand number even when you rank off-page, read brandShareOfVoice from the aggregated endpoint — it's resolved server-side against the full mention list.

curl "https://api.getmint.ai/api/domains/$DOMAIN_ID/topics/$TOPIC_ID/visibility/mentions?limit=20&page=1" \
  -H "X-API-Key: $MINT_API_KEY"

Response (trimmed; full shape in API reference):

{
  "mentions": [
    { "mention": "Nike", "count": 312, "percentage": 34 },
    { "mention": "Adidas", "count": 238, "percentage": 26 },
    { "mention": "Puma", "count": 174, "percentage": 19 }
  ],
  "mentionsDetailed": [
    { "mention": "Nike", "modelId": "gpt-5", "count": 184 },
    { "mention": "Nike", "modelId": "claude-sonnet-4", "count": 128 }
  ],
  "total": 156,
  "page": 1,
  "limit": 20,
  "totalPages": 8
}

Query parameters

Prop

Type

Variations

Change	Effect
Read `topMentions` from `.../visibility/aggregated`	Top 10 in the call you already make for the score — zero extra requests
`models=gpt-5`	Leaderboard scoped to a single model
`startDate`/`endDate`	Leaderboard for a specific window
`page=2`	Next page of mentions; rank continues via `(page - 1) * limit`

Common follow-ups

Compare to competitors

Turn the leaderboard into a per-competitor time series.

Ask an agent instead

The get_topic_mentions MCP tool returns this leaderboard, agent-ready.

{ "mentions": [ { "mention": "Nike", "count": 312, "percentage": 34 }, { "mention": "Adidas", "count": 238, "percentage": 26 }, { "mention": "Puma", "count": 174, "percentage": 19 } ], "mentionsDetailed": [ { "mention": "Nike", "modelId": "gpt-5", "count": 184 }, { "mention": "Nike", "modelId": "claude-sonnet-4", "count": 128 } ], "total": 156, "page": 1, "limit": 20, "totalPages": 8 }

Change

Effect

Read topMentions from .../visibility/aggregated

Top 10 in the call you already make for the score — zero extra requests

models=gpt-5

Leaderboard scoped to a single model

startDate/endDate

Leaderboard for a specific window

page=2

Next page of mentions; rank continues via (page - 1) * limit