# ChatGPT Scraper | Answers, Citations & Brand Mentions (`alexmorain/chatgpt-search-scraper`) Actor

Send prompts to ChatGPT search and get the full answer, every cited source, the brands it names and their rank, and the web searches it ran. Ask each prompt up to 5 times for citation rates, track your domains and brands, and see what changed since your last run.

- **URL**: https://apify.com/alexmorain/chatgpt-search-scraper.md
- **Developed by:** [Alexis Morain](https://apify.com/alexmorain) (community)
- **Categories:** AI, SEO tools, Marketing
- **Features:** Limited permissions
- **Stats:** 4 total users, 4 monthly users, 100.0% runs succeeded, 0 bookmarks
- **User rating**: No ratings yet

## Pricing

from $2.00 / 1,000 prompt checked, 1 or 2 captures

This Actor is paid per event. You are not charged for the Apify platform usage, but only a fixed price for specific events.
Since this Actor supports Apify Store discounts, the price gets lower the higher subscription plan you have.

Learn more: https://docs.apify.com/actors/running/actors-in-store.md#pay-per-event

## What's an Apify Actor?

An Actor is a serverless cloud program that runs on the Apify platform. It has two run modes.
In Batch mode, an Actor accepts a well-defined JSON input, performs an action which can take anything from a few seconds to a few hours,
and optionally produces a well-defined JSON output, datasets with results, or files in key-value store.
In Standby mode, an Actor provides a web server which can be used as a website, API, or an MCP server.

Apify vocabulary and the platform model are defined once, in the agent quickstart at https://apify.com/agents.md.

## How to integrate an Actor?

If asked about integration, you help developers integrate Actors into their projects.
You adapt to their stack and deliver integrations that are safe, well-documented, and production-ready.

Do not guess an integration path. Every one of them is in the agent quickstart at https://apify.com/agents.md: the Apify MCP server, Agent Skills with the Apify CLI, the JavaScript and Python clients, the REST API, and the account-free path for an agent with no human to sign in. It also carries the rule on stating cost before the first paid run.

For examples already wired to this Actor's own input schema, see the [API](#api) section below.

Each client library has reference documentation the quickstart does not restate: [JavaScript/TypeScript](https://docs.apify.com/api/client/js/docs.md) (`npm install apify-client`) and [Python](https://docs.apify.com/api/client/python/docs.md) (`pip install apify-client`).

# README

### What does ChatGPT Search Scraper do?

ChatGPT Search Scraper sends your prompts to **ChatGPT with web search** and returns what it answers: the full answer, every source it cites, the brands it names, the follow-up searches it ran, and the search results it read. You give it prompts like `best crm software for startups`. For each one you get a citation rate per domain and a mention rate and rank per brand, measured over 1, 3 or 5 answers to the same prompt.

Use it to see whether ChatGPT recommends your brand, which sites it trusts in your market, and how that moves from week to week. It works as a ChatGPT visibility API for SEO, GEO (generative engine optimization) and brand reporting. No ChatGPT account, no login, no browser on your side.

It covers ChatGPT only. For Google AI Overviews, Google AI Mode, Perplexity or Gemini, see our other Actors.

### Why use ChatGPT Search Scraper?

ChatGPT gives a different answer each time. We asked `best crm software for startups` three times in a row on November 9, 2026, from the United States. The three answers cited 12 domains in total, and only 7 of them appeared in all three. A single answer would have reported 5 domains as "cited" or "not cited" by luck. This Actor asks each prompt 1, 3 or 5 times and gives every domain and brand a rate: 0.67 means it showed up in 2 answers out of 3.

The second reason is brand names. ChatGPT tags the brands it recommends, but the tag label is not always the name. In one of our three answers, all 8 tags read like "Customer relationship management platform" or "Sales CRM platform" instead of HubSpot or Pipedrive. A scraper that reads the label reports zero brands without any error. This Actor reads the brand name ChatGPT actually wrote, and checks your own brand names against the answer text.

What you get on top:

- Brand rank: the order in which each brand first appears in the answer, so you see whether you come first or fifth.
- Tracked domains: citation rate and best position in the source list for your site and your competitors. A tracked domain also counts its subdomains.
- Fan-out queries: the web searches ChatGPT ran before answering. They tell you which pages to write.
- Search results: the pages ChatGPT read, cited or not.
- Compare mode: domains and brands gained or lost since the previous run, with old and new rates.
- Clean URLs: the `utm_source=chatgpt.com` tag is removed.
- 191 countries and 128 languages.
- Scheduling, API access, webhooks and integrations (Make, Zapier, n8n, Google Sheets) from the Apify platform.

### What data can ChatGPT Search Scraper extract?

One result per prompt:

| Field | Type | Description |
|---|---|---|
| `prompt` | string | The prompt as you wrote it |
| `model` | string | ChatGPT model that answered |
| `status` | string | `answered` or `failed` |
| `capturesSucceeded` | integer | Answers received. Every rate is divided by this number |
| `answerText` | string | Full answer in Markdown (the longest of the captures) |
| `citedDomains` | array | Each cited domain with its captures, rate and best position |
| `sources` | array | Each cited page: URL, title, rate, best position |
| `brandMentions` | array | Every brand named, with its rate and best rank in the answer |
| `trackedDomains` | array | Your domains: citation rate and best position |
| `trackedBrands` | array | Your brands: mention rate and best rank |
| `fanOutQueries` | array | Web searches ChatGPT ran for this prompt |
| `searchResults` | array | Pages ChatGPT read, with their position |
| `changes` | object | Compare mode: gained and lost domains and brands, tracked rates before and after |

### How to scrape ChatGPT answers

1. Open the Actor and type your prompts, one per line.
2. Pick the country and language, for example `us` and `en`, or `fr` and `fr`.
3. Keep 1 capture for a snapshot, or pick 3 or 5 to get rates you can compare between runs.
4. Add your domain and competitor domains in Domains to track, and brand names in Brands to track.
5. Switch on Compare with previous run if you plan to run it on a schedule.
6. Press Start, then download the results as JSON, CSV or Excel, or read them through the API.

Answers come back in batches, not instantly: the prompts are queued and ChatGPT answers them in the background. In our tests a run of 200 prompts with 1 capture finished in under 4 minutes.

### How much does it cost to scrape ChatGPT?

You pay per prompt answered, plus a start fee of $0.001 per run. The price grows with the number of captures, because each capture is a real ChatGPT answer:

| Captures per prompt | Free plan | Bronze | Silver | Gold and above |
|---|---|---|---|---|
| 1 (snapshot) | $0.003 | $0.003 | $0.0025 | $0.002 |
| 3 (rates) | $0.009 | $0.009 | $0.0075 | $0.006 |
| 5 (most reliable) | $0.015 | $0.015 | $0.0125 | $0.010 |

1,000 prompts with 1 capture cost $3 on the Free plan and $2 on Gold. Platform usage is included.

A prompt is charged only when at least one answer came back. If some captures fail, the prompt is charged at the level it reached: 3 asked and 2 received is charged as 1 capture. A prompt with no answer comes back with `status: failed` and is not charged. The run stops before it would go over the maximum cost you set.

Free Apify plan: up to 0.15 USD of DataForSEO data per day; paid plans unlimited.

### Input

See the Input tab for every option. The only required field is the list of prompts, up to 1,000 per run and 2,000 characters each.

```json
{
    "prompts": ["best crm software for startups", "best email marketing tool for ecommerce"],
    "countryCode": "us",
    "languageCode": "en",
    "capturesPerPrompt": "3",
    "trackDomains": ["hubspot.com", "zoho.com"],
    "trackBrands": ["HubSpot", "Salesforce"],
    "compareWithPreviousRun": true,
    "historyKey": "crm-weekly"
}
```

### Output

You can download the dataset in various formats such as JSON, HTML, CSV, or Excel. Example from a real run, shortened:

```json
{
    "prompt": "best crm software for startups",
    "engine": "chatgpt",
    "model": "gpt-6",
    "countryCode": "us",
    "languageCode": "en",
    "status": "answered",
    "capturesRequested": 3,
    "capturesSucceeded": 3,
    "answerText": "# Best CRM software for startups in 2026\n\nThe best overall CRM for most startups is **HubSpot**...",
    "citedDomains": [
        {"domain": "hubspot.com", "captures": 3, "rate": 1.0, "bestPosition": 1},
        {"domain": "attio.com", "captures": 3, "rate": 1.0, "bestPosition": 2},
        {"domain": "zoho.com", "captures": 3, "rate": 1.0, "bestPosition": 3}
    ],
    "brandMentions": [
        {"brand": "HubSpot", "captures": 3, "rate": 1.0, "bestRank": 1},
        {"brand": "Pipedrive", "captures": 3, "rate": 1.0, "bestRank": 2},
        {"brand": "Freshsales", "captures": 1, "rate": 0.33, "bestRank": 7}
    ],
    "trackedDomains": [
        {"domain": "hubspot.com", "citedCaptures": 3, "citationRate": 1.0, "bestPosition": 1}
    ],
    "trackedBrands": [
        {"brand": "Salesforce", "mentionedCaptures": 3, "mentionRate": 1.0, "bestRank": 4}
    ],
    "fanOutQueries": [
        "best CRM for startups 2026 HubSpot Pipedrive Zoho Salesforce pricing free plans"
    ],
    "changes": null
}
```

`changes` is null on the first run of a prompt: that run becomes the baseline for the next one.

### Tips

- Use 3 captures for weekly monitoring. With 1 capture, a domain that ChatGPT cites half the time looks gained one week and lost the next.
- Give each client or project its own History name, so compare mode keeps their histories apart.
- Write prompts the way people ask ChatGPT, as full questions, and in the language you pick.
- Brand names are matched as whole words, case ignored. Add the spellings people use (`monday.com` and `Monday`).

### FAQ, limits and support

#### Which ChatGPT does it ask?

ChatGPT on chatgpt.com, with web search, the way it answers in the web interface. The `model` field tells you which model answered. Answers are not personalized: no account, no memory, no history.

#### Why do the cited sources change between captures?

ChatGPT searches the web again for each answer and does not always pick the same pages. That is why rates over 3 or 5 captures say more than one answer.

#### Is it legal?

Our Actors are ethical and do not extract any private user data, such as email addresses, gender, or location. They only extract what the user has chosen to share publicly. We therefore believe that our Actors, when used for ethical purposes by Apify users, are safe. However, you should be aware that your results could contain personal data. Personal data is protected by the GDPR in the European Union and by other regulations around the world. You should not scrape personal data unless you have a legitimate reason to do so. If you're unsure whether your reason is legitimate, consult your lawyers.

#### Something is wrong or missing?

Open an issue in the Issues tab. To call the Actor from your own code, see the API tab.

### Guide and use cases

Use cases for outbound and automation, the output fields, pricing and common questions are on the [ChatGPT Scraper page](https://alexis.morain.fr/en/scrapers/chatgpt-search-scraper/) of my site. The [AI search visibility family page](https://alexis.morain.fr/en/scrapers/families/ai-search/) puts it side by side with the other scrapers of the same family.

# Actor input Schema

## `prompts` (type: `array`):

What people ask ChatGPT, one prompt per line: <code>best crm for a startup</code>. From 1 to 1,000 per run, up to 2,000 characters each (longer prompts are refused, never cut). Duplicates are removed.

## `countryCode` (type: `string`):

Where the prompt is asked from. The answer and its sources change with the country.

## `languageCode` (type: `string`):

Language of the prompt and of the answer. Match it to the language of your prompts.

## `capturesPerPrompt` (type: `string`):

How many times each prompt is asked. ChatGPT answers differently each time, so one capture is a snapshot: with 3 or 5, every domain and brand gets a rate (0.67 = present in 2 captures out of 3). The price per prompt grows with the captures that succeed.

## `trackDomains` (type: `array`):

Your website and your competitors, one per line: <code>hubspot.com</code> (50 at most). A full URL works too, only the domain is kept, and subdomains count. Each prompt gets a citation rate and a best position per domain.

## `trackBrands` (type: `array`):

Brand or product names to look for in the answer text, one per line (50 at most): <code>HubSpot</code>, <code>monday.com</code>. Whole word, case ignored. Each prompt gets a mention rate and a rank per brand.

## `compareWithPreviousRun` (type: `boolean`):

Keep this run's results and compare the next one with them. Each prompt then gets a <code>changes</code> object: domains and brands gained or lost (present in at least half the captures), and the old and new rates of what you track. The first run saves the baseline. Use it with a schedule for weekly monitoring.

## `historyKey` (type: `string`):

Name of the saved history used by <b>Compare with previous run</b>. Give each client or project its own name to keep them separate. Prompts are kept apart by country and language.

## `includeAnswerText` (type: `boolean`):

Add the answer (the longest of the captures) to each result.

## `includeSources` (type: `boolean`):

Add every cited source (title, URL, rate) and, where the engine gives them, its search results. Turn it off to keep only the per-domain summary.

## `webhookUrl` (type: `string`):

Optional. A public URL that receives a short JSON summary when the run ends (Slack, Discord, Make, Zapier, n8n). It is never written to the log or the results.

## Actor input object example

```json
{
  "prompts": [
    "best crm for startups"
  ],
  "countryCode": "us",
  "languageCode": "en",
  "capturesPerPrompt": "1",
  "compareWithPreviousRun": false,
  "historyKey": "default",
  "includeAnswerText": true,
  "includeSources": true
}
```

# Actor output Schema

## `results` (type: `string`):

All rows in the default dataset.

## `summary` (type: `string`):

Prompts answered and failed, events charged, stop reason.

# API

You can run this Actor programmatically using our API. Below are code examples in JavaScript, Python, and CLI, as well as the OpenAPI specification and MCP server setup.

## JavaScript example

```javascript
import { ApifyClient } from 'apify-client';

// Initialize the ApifyClient with your Apify API token
// Replace the '<YOUR_API_TOKEN>' with your token
const client = new ApifyClient({
    token: '<YOUR_API_TOKEN>',
});

// Prepare Actor input
const input = {
    "prompts": [
        "best crm for startups"
    ]
};

// Run the Actor and wait for it to finish
const run = await client.actor("alexmorain/chatgpt-search-scraper").call(input);

// Fetch and print Actor results from the run's dataset (if any)
console.log('Results from dataset');
console.log(`💾 Check your data here: https://console.apify.com/storage/datasets/${run.defaultDatasetId}`);
const { items } = await client.dataset(run.defaultDatasetId).listItems();
items.forEach((item) => {
    console.dir(item);
});

// 📚 Want to learn more 📖? Go to → https://docs.apify.com/api/client/js/docs

```

## Python example

```python
from apify_client import ApifyClient

# Initialize the ApifyClient with your Apify API token
# Replace '<YOUR_API_TOKEN>' with your token.
client = ApifyClient("<YOUR_API_TOKEN>")

# Prepare the Actor input
run_input = { "prompts": ["best crm for startups"] }

# Run the Actor and wait for it to finish
run = client.actor("alexmorain/chatgpt-search-scraper").call(run_input=run_input)

# Fetch and print Actor results from the run's dataset (if there are any)
print(f"💾 Check your data here: https://console.apify.com/storage/datasets/{run.default_dataset_id}")
for item in client.dataset(run.default_dataset_id).iterate_items():
    print(item)

# 📚 Want to learn more 📖? Go to → https://docs.apify.com/api/client/python/docs/quick-start

```

## CLI example

```bash
echo '{
  "prompts": [
    "best crm for startups"
  ]
}' |
apify call alexmorain/chatgpt-search-scraper --silent --output-dataset

```

## MCP server setup

```json
{
    "mcpServers": {
        "apify": {
            "type": "http",
            "url": "https://mcp.apify.com/?tools=fetch-actor-details,alexmorain/chatgpt-search-scraper"
        }
    }
}
```

The hosted server signs you in with OAuth on first connect, so no API token belongs in this config. Clients without OAuth support can send an `Authorization: Bearer <APIFY_API_TOKEN>` header instead, using a token from API & Integrations in Apify Console (https://console.apify.com/settings/integrations).

## OpenAPI specification

Download the OpenAPI definition: https://api.apify.com/v2/actors/gD9LiwMu42daOG7kr/builds/uvGRGXxfsu3gaGdGl/openapi.json
