domains | string or array of string | No | - | Domains that must appear in the AI answer, as a list or newline/comma-separated text (bare domains such as ahrefs.com). Use this to track a brand’s website being cited or linked. At least one domain or keyword is required; up to 10 targets in total. |
keywords | string or array of string | No | - | Words or phrases that must appear in the AI answer, question or named brands, e.g. a brand name or a product category. List or newline/comma-separated text. All included domains and keywords must match together (AND). |
excluded_domains | string or array of string | No | - | Domains that must NOT appear in the matched answers, e.g. wikipedia.org. |
excluded_keywords | string or array of string | No | - | Words or phrases that must NOT appear in the matched answers. |
domain_scope | string or SelectableOption | No | any | Where a domain has to appear: ‘any’ (default), ‘sources’ (cited in the answer) or ‘search_results’ (retrieved by the web search behind the answer; ChatGPT only). |
keyword_scope | string or SelectableOption | No | any | Where a keyword has to appear: ‘any’ (default), ‘question’ (the user prompt), ‘answer’ (the AI answer text), ‘brand_entities’ (brands named in the answer) or ‘fan_out_queries’ (the follow-up searches the model ran). |
keyword_match_type | string or SelectableOption | No | word_match | ’word_match’ (default) matches whole words, so ‘light’ also matches ‘light bulb’; ‘partial_match’ matches substrings, so ‘light’ also matches ‘lighting’. |
include_subdomains | boolean | No | True | Also match subdomains of the target domains, which includes the www. host most sites are cited under. Defaults to true; turn it off to count only the bare domain. |
platform | string or SelectableOption | No | all | AI platform: ‘all’ (default), ‘google’ (Google AI Overviews and AI Mode) or ‘chat_gpt’. ChatGPT data exists for the United States in English only; every other location and language is Google only. |
location | string or SelectableOption | No | 2840 | Country or location: a two-letter country code (us, gb, de…), a DataForSEO location code (2840 = United States) or a full location name. Defaults to the United States. Only locations listed by DataForSEO for LLM Mentions have data. |
language | string or SelectableOption | No | en | Language code of the AI answers, e.g. en, de, es, fr, nl. Defaults to en. |
filters | array of FilterItem | No | - | Up to 8 filters on ai_search_volume, model_name, platform, location_code, language_code, is_web_search_based, first_response_at or last_response_at, e.g. ai_search_volume GREATER_THAN 100. |
orders | array of OrderItem | No | - | Sort by up to 3 of the filterable fields, e.g. ai_search_volume DESC or last_response_at DESC. |
limit | integer | No | 20 | Maximum number of AI answers to return (1 to 1000). Defaults to 20. Each answer is a full markdown response with its sources, so keep this small; DataForSEO also bills per row on top of the request fee. |
offset | integer | No | - | Number of answers to skip, for paging. |
fields | string or array of string | No | - | Result columns to keep: platform, model_name, question, answer, ai_search_volume, monthly_searches, sources, search_results, brand_entities, fan_out_queries, is_web_search_based, first_response_at, last_response_at, location_code, language_code. Leave empty for the default set listed in the output description. |