LLM Mentions

Find where your brand, domain, or keyword is mentioned across ChatGPT and Google AI Overviews. Pre-indexed mentions corpus with AI search volume.

POST/v1/seo/ai/llm-mentions

$0.20/call

Search a pre-indexed corpus of LLM answers for any domain or keyword, across ChatGPT and Google AI Overviews. Returns the full AI answer, the cited sources, AI search volume, mentioned brand entities, and fan-out queries the LLM derived from the original question.

Use this to monitor how often your domain shows up as a source in AI-generated answers, which competitors get cited for your keywords, and which prompts surface your content in ChatGPT / Google AI Overview.

Usage

const res = await fetch('https://api.yepapi.com/v1/seo/ai/llm-mentions', {
  method: 'POST',
  headers: {
    'x-api-key': 'YOUR_API_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    target: [
      { domain: 'yepapi.com', search_filter: 'include' },
      { keyword: 'seo api', search_scope: ['answer'] },
    ],
    platform: 'google',
    location: 'us',
    language: 'en',
    limit: 100,
    filters: [['ai_search_volume', '>', 100]],
    order_by: ['ai_search_volume,desc'],
  }),
});
const { data } = await res.json();
console.log(data);

curl -X POST https://api.yepapi.com/v1/seo/ai/llm-mentions \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "target": [{ "domain": "yepapi.com" }],
    "platform": "google",
    "limit": 50
  }'

Request Body

Parameter	Type	Required	Description	Default
`target`	`array`	Yes	1–10 entries. Each entry is `{ domain }` or `{ keyword }` with optional `search_filter`, `search_scope`, `include_subdomains` (domains), or `match_type` (keywords).	—
`platform`	`string`	No	`"google"` (AI Overview, multi-locale) or `"chat_gpt"` (US/English only).	`"google"`
`location`	`string`	No	Country code (e.g. `"us"`, `"gb"`, `"de"`).	`"us"`
`location_code`	`integer`	No	Raw DataForSEO location code. Overrides `location`.	—
`language`	`string`	No	Language code (e.g. `"en"`).	`"en"`
`filters`	`array`	No	Up to 8 filters. Each is `[field, operator, value]`. Operators: `=`, `<>`, `<`, `>`, `<=`, `>=`, `in`, `not_in`, `like`, `not_like`, `ilike`, `not_ilike`, `match`, `not_match`. Combine with `"and"` / `"or"`.	—
`order_by`	`array<string>`	No	Up to 3 sort rules, e.g. `["ai_search_volume,desc"]`.	—
`offset`	`integer`	No	Skip results (0–9000). Use `search_after_token` for deeper pagination.	`0`
`search_after_token`	`string`	No	Pagination cursor returned by a previous response.	—
`limit`	`integer`	No	Items returned (1–1000).	`100`
`tag`	`string`	No	User-defined task tag (≤255 chars).	—

Target entry shape

Domain entry

Field	Type	Description
`domain`	`string`	Bare domain (no scheme, no `www.`). Max 63 chars.
`search_filter`	`string`	`"include"` (default) or `"exclude"`.
`search_scope`	`array<string>`	Any of `"any"`, `"sources"`, `"search_results"`. Default `["any"]`.
`include_subdomains`	`boolean`	Include subdomain matches. Default `false`.

Keyword entry

Field	Type	Description
`keyword`	`string`	Up to 250 chars.
`search_filter`	`string`	`"include"` (default) or `"exclude"`.
`search_scope`	`array<string>`	Any of `"any"`, `"question"`, `"answer"`, `"brand_entities"`, `"fan_out_queries"`.
`match_type`	`string`	`"word_match"` (default) or `"partial_match"`.

Response

{
  "ok": true,
  "data": {
    "platform": "google",
    "totalCount": 1842,
    "itemsCount": 2,
    "currentOffset": 0,
    "searchAfterToken": "eyJzZWFyY2hfYWZ0ZXIiOlsx...",
    "mentions": [
      {
        "platform": "google",
        "modelName": "google_ai_overview",
        "locationCode": 2840,
        "languageCode": "en",
        "question": "what is the best seo api",
        "answer": "Several SEO APIs stand out in 2026, including YepAPI for keyword research and SERP data, DataForSEO for enterprise-scale crawling...",
        "sources": [
          {
            "position": 1,
            "title": "Keywords API — YepAPI",
            "url": "https://yepapi.com/seo-api/keywords",
            "domain": "yepapi.com",
            "sourceName": "YepAPI",
            "snippet": "Bulk keyword metrics: search volume, CPC, competition, difficulty, intent...",
            "thumbnail": null,
            "markdown": null,
            "publicationDate": null
          }
        ],
        "searchResults": null,
        "aiSearchVolume": 1200,
        "monthlySearches": [
          { "year": 2026, "month": 4, "searchVolume": 1200 },
          { "year": 2026, "month": 3, "searchVolume": 980 }
        ],
        "firstResponseAt": "2026-02-11 08:14:00 +00:00",
        "lastResponseAt": "2026-05-15 22:41:00 +00:00",
        "brandEntities": [
          { "position": 1, "title": "Developer-first SEO + AI API", "category": "software" }
        ],
        "fanOutQueries": null
      }
    ]
  }
}

Response Fields

Field	Type	Description
`ok`	`boolean`	Whether the request succeeded.
`data.platform`	`string`	Platform queried (`google` or `chatGpt`).
`data.totalCount`	`integer`	Total matching records in the full index.
`data.itemsCount`	`integer`	Items returned in this page.
`data.currentOffset`	`integer`	Offset used for this page.
`data.searchAfterToken`	`string`	Cursor for retrieving the next page beyond offset 9000.
`data.mentions[]`	`array`	Mention records (see fields below).
`data.mentions[].platform`	`string`	`google` or `chatGpt`.
`data.mentions[].modelName`	`string`	AI model identifier (e.g. `google_ai_overview`).
`data.mentions[].question`	`string`	The user prompt that triggered the answer.
`data.mentions[].answer`	`string`	Full LLM response in markdown.
`data.mentions[].sources[]`	`array`	Cited sources — `position`, `title`, `url`, `domain`, `sourceName`, `snippet`, `thumbnail`, `markdown`, `publicationDate`.
`data.mentions[].searchResults[]`	`array`	Web results the LLM retrieved while answering.
`data.mentions[].aiSearchVolume`	`integer`	Current AI search volume for the question.
`data.mentions[].monthlySearches[]`	`array`	Historical monthly search volumes.
`data.mentions[].firstResponseAt`	`string`	UTC timestamp when the answer was first recorded.
`data.mentions[].lastResponseAt`	`string`	UTC timestamp of the most recent re-index.
`data.mentions[].brandEntities[]`	`array`	Brand entities mentioned in the answer — `position`, `title`, `category`.
`data.mentions[].fanOutQueries[]`	`array`	Related queries the LLM derived from the original question.

Pricing

$0.20 base + $0.002 per row returned. A request that returns 100 mentions costs $0.40. The minimum balance gate is $0.20.

Limits

Up to 10 target entries per request.
Up to 8 filters and 3 sort rules.
offset capped at 9000 — use searchAfterToken for deeper pagination.
limit capped at 1000.
platform: "chat_gpt" is US/English only (location_code 2840, language en).

Under the Hood

Backed by an indexed corpus of ~253M LLM prompts — far cheaper and faster than re-scraping the LLMs live. Results are cached for 6 hours.