Track how often ChatGPT and Google AI Overviews mention your brand
Track brand and domain mentions across ChatGPT and Google AI Overviews using the vibiz CLI, HTTP API, or @vibiz/app.
On this page 8 sections
You can track brand and domain mentions across ChatGPT and Google AI Overviews using the seo.ai_mentions action in the thelastapi platform. This single action scans prompts and answers to return the specific user questions asked, relevant answer excerpts, and cited URLs. You can query these mentions via the vibiz CLI, a direct HTTP POST request, or the @vibiz/app SDK.
Search engines and AI assistants increasingly answer questions directly instead of showing a list of links. When an answer refers to your product or cites your site, it shapes how buyers discover you. Traditional rank tracking does not show these AI-generated responses. Using web data actions, you can monitor where your brand appears across conversational models and AI search summaries.
How AI mention tracking works
The seo.ai_mentions action monitors conversational interfaces to discover prompts where a target brand or domain is referenced.
The action accepts either a domain (such as example.com) or a brand name (such as "Acme"). It queries ChatGPT and Google's AI Overviews, then returns structured records containing:
- The platform where the mention occurred (
chatgptorgoogle). - The user question or prompt.
- An excerpt of the generated answer where the brand or site appeared.
- The cited URL, when available.
- Summary totals broken down by platform.
Because checking multiple AI platforms requires querying live models, an seo.ai_mentions run can take up to two minutes to complete.
Checking mentions with the CLI
The quickest way to check your presence is using the vibiz seo ai command in your terminal. You can pass a domain name or brand string:
vibiz seo ai example.comBy default, the command searches both ChatGPT and Google AI Overviews. If you only want to inspect a single service, use the --platform flag:
vibiz seo ai "Acme" --platform chatgptTo parse the output inside your own scripts, append the --json flag:
vibiz seo ai example.com --platform google --jsonThis returns a JSON object containing your target query, total mention counts, and an array of individual citation objects.
Running mention checks via the API
You can call the action directly over HTTP from any language. Make a POST request to the /v1/actions/seo.ai_mentions endpoint with your API key in the authorization header:
curl https://thelastapi.dev/v1/actions/seo.ai_mentions \
-H "Authorization: Bearer $VIBIZ_API_KEY" \
-H "content-type: application/json" \
-d '{
"input": {
"target": "example.com",
"platform": "all",
"limit": 20
}
}'The response includes the run metadata, processing cost, and the structured mention data:
{
"object": "action_result",
"action": "seo.ai_mentions",
"output": {
"target": "example.com",
"kind": "domain",
"total": 3,
"by_platform": {
"chatgpt": 1,
"google": 2
},
"mentions": [
{
"platform": "google",
"question": "what are the best tools for web scraping?",
"answer_excerpt": "...for developers needing clean extraction, example.com offers structured endpoints...",
"url": "https://example.com/blog"
}
]
},
"cost_micro_usd": 40000,
"run_id": "arun_12345",
"console_url": "https://thelastapi.dev/web-data?action=seo.ai_mentions&run=arun_12345"
}The action can also be invoked inside an application using @vibiz/app:
import { vibiz } from '@vibiz/app';
const result = await vibiz.actions.run('seo.ai_mentions', {
target: 'example.com',
platform: 'all',
limit: 20
});
console.log(`Found ${result.output.total} mentions`);
for (const mention of result.output.mentions) {
console.log(`[${mention.platform}] ${mention.question}`);
}Input parameters and limits
The seo.ai_mentions action takes three primary parameters:
| Field | Type | Default | Details |
|---|---|---|---|
target | string | required | A domain name (e.g., example.com) or a brand string (e.g., "Acme"). |
platform | string | "all" | Supported values are "all", "google", or "chatgpt". |
limit | number | 20 | Mentions per platform to return. Accepts values between 1 and 100. |
Because this action queries external platforms directly, calls bound between 60 and 150 seconds. If an operation exceeds its limit, it fails with a 503 provider_unavailable status and details.reason: "timeout". If you encounter timeouts, ask for fewer results by setting a smaller limit.
Inspecting past runs
Every action executed through the CLI, HTTP API, or SDK is logged automatically. You can review past runs from the console under Web data, or using the CLI:
# List past AI mention runs
vibiz runs --action seo.ai_mentions --since 7d
# Inspect a specific run by id
vibiz runs arun_12345 --jsonYou can also fetch past runs programmatically:
GET /v1/actions/runs: lists past runs with metadata.GET /v1/actions/runs/:id: retrieves the full inputs, outputs, timestamps, and cost for a specific run.
Using the run history prevents duplicate queries and helps track how brand citations change over time.
Pricing and spend ceilings
Web data calls are billed by resource usage plus a 30% margin, deducted from account credits:
- AI mention tracking costs roughly 2 cents per platform queried (about 4 cents total when
platformis set to"all"). - Every call reports its exact base cost in the response via
cost_micro_usd. - Spend logs are viewable through
vibiz usageor inside the console billing overview. - All web data calls count toward your account's daily spend ceiling ($5 per day by default, which can be modified by the account owner).
If your account hits the ceiling or runs out of credits, calls return 429 rate_limited or 402 payment_required.
Combining AI mentions with traditional SEO data
Monitoring AI citations is one part of understanding online visibility. You can pair seo.ai_mentions with other actions available under web data actions:
- SERP features: Use the SERP API via
vibiz seo serp "<keyword>"to check if Google displays an AI Overview block (ai_overview) alongside standard organic results. - Page content audits: Run
vibiz seo audit <url>to inspect page load speeds, title tags, and missing headers to ensure source pages can be indexed cleanly. - Page extraction: Use markdown scraping via
vibiz scrape <url>to see the exact text content an automated crawler extracts from your landing pages. - Automated engine submission: Connect your domains through the search setup to push verified URLs into Google Search Console and Bing via IndexNow after every deploy.
To begin tracking mentions, create your account at thelastapi.dev/start to access the API and CLI tools.
Frequently asked questions
How long does an seo.ai_mentions call take to run?
The action queries conversational interfaces in real time, so runs can take up to two minutes to finish. If an operation takes longer than 150 seconds, it aborts with a 503 timeout error.
Can I check mentions for a brand name rather than a web domain?
Yes, the target input accepts either a domain like example.com or a brand name like "Acme". The action automatically detects whether the target is a brand or a domain and sets the kind output field accordingly.
What platforms are supported by seo.ai_mentions?
The action currently tracks mentions across ChatGPT and Google AI Overviews. You can monitor both together by setting the platform parameter to all, or limit queries to chatgpt or google.
What happens if an AI mention run fails?
If a call fails before starting, no usage is charged. However, if an action starts processing and fails mid-run due to an external provider timeout, it bills what it used up to that point.