The AI mentions API. Every answer that named you, with the source.
Each mention of a brand across ChatGPT, Google AI Mode, Google AI Overviews, Gemini, Claude, Perplexity, Grok and DeepSeek, as a record: which engine, which prompt, where in the answer, how it was described, and which sources the engine cited to say it.
8
engines watched for brand and competitor mentions
1 record
per mention, with the full answer text and the citing sources
$0.08
per answer on every engine, from prepaid credits that do not expire
What it returns
Mentions with the receipts attached.
A mention is only useful with its context. Each record carries the engine, the prompt, whether the brand was merely mentioned or actually recommended, its position among the named options, the sentiment of the framing, an excerpt of the answer, and the list of sources the engine cited, each classified as brand-owned, competitor, review platform, directory, publisher or community.
- Mentioned versus recommended. The two are separated in the payload. A brand named as a caution is not a win, and the data should not count it as one.
- Classified citations. Every cited domain comes typed, so an integration can tell a review platform from a competitor's blog without its own lookup table.
- Competitors too. Ask for the mentions of any brand in the category, not just your own. Share of the answer is derived from the same records.
Endpoints
Mentions, citations and your own prompts.
Read what the scheduled runs found, or submit prompts and get every enriched answer back.
Every answer that named the brand: engine, prompt, position, sentiment, excerpt and the competitors named alongside.
The sources behind the answers: every cited page with its domain, page type and position, so you can see which platforms shape the category.
Submit a prompt set and a brand. Each answer comes back enriched, billed per answer.
Task status per engine and the settled cost.
Your tasks with status and spend.
Build with it
What teams build on mentions data.
Brand monitoring for AI answers
Extend a social listening or brand monitoring product to the answers ChatGPT and Perplexity give, with the same record shape as a social mention.
Source intelligence
See which review platforms, directories and publishers the engines lean on in a category, then work on the ones the brand is absent from.
PR measurement
Tie a placement to the answers that started citing it. The citations endpoint returns the domain and the date it first appeared.
Competitive research
Pull every competitor mention in the category and see how each is described, and by which engine.
Why it holds up
Why mention data from Honeyb holds up.
Detection reads the answer
Mentions are found in the generated text and the cited sources, not by keyword matching a page title. Variants, product names and misspellings are handled.
Repeated runs
A mention that appears in one run out of thirty is reported as such. The record set covers every run, and aggregates carry the sample count.
Grounded engines with sources
Where the engine searched the web before answering, the sources it used are captured and classified. That is where the leverage is.
Pricing
Billed per answer.
Prompts submitted through the task API are billed per answer: $0.08 on every engine, from a credit balance of prepaid credits that do not expire. No subscription.
FAQ
Frequently asked questions.
More from the Honeyb API
AI Visibility API
LLM visibility API: run your prompts across eight AI engines and get back whether a brand was named, where, with what sentiment and which pages were cited. $0.08 per answer, no subscription.
ExploreSERP API
SERP API: structured Google and Bing results with organic, paid, People also ask, local pack and the full AI Overview with its cited sources, for any location and device. Live or queued.
ExploreRank Tracking API
Rank tracking API: daily Google and Bing positions by keyword, location and device, with SERP features and AI answer visibility in the same response. Usage-based, no subscription.
ExploreRelated on Honeyb: Citation tracking, Brand mentions guide, Sentiment analysis.
Free to start
Get recommended by AI search models.
Run a free AI search visibility check, then let the Honeyb agent do the work that gets you into the answers.
ChatGPT
Claude
Gemini