Pricing
Free and open source under the MIT licence. Sampling ChatGPT requires your own ChatGPT account, whose cost is separate and not covered on the page.
Learn moreOpen source Python tool that logs AI answers and the sources they cite

frankvitetta is an AI aggregators tool. AI Scraper is an MIT-licensed Python tool that runs a CSV list of prompts through Perplexity or ChatGPT in a real Chrome browser and saves each answer with its cited sources, for small-sample AI visibility research.
AI Scraper is a small open source Python tool for people who want to see what answer engines say about a chosen set of questions. It takes a list of prompts, feeds them one at a time to Perplexity or ChatGPT, and writes down both the answer text and the addresses of the sources each answer cited. The project comes from LLM Scout, is published on GitHub, and is released under the MIT licence. In practice the workflow is simple. Prompts go into a CSV file with one prompt per row. The tool launches a real Chrome window through Selenium and types the prompts in sequence, pacing its requests instead of firing them in a burst. Results are saved to a CSV file, with JSONL as an optional output format. Perplexity can be sampled directly, while ChatGPT requires the user to log in manually inside the browser window the tool opens, so a personal ChatGPT account is needed for that engine. A cap on the number of prompts makes it easy to start with a small trial run before committing to a larger list. The audience is marketers, SEO practitioners and researchers who are interested in AI visibility, meaning which brands and pages show up as citations in generated answers. The author is explicit that the output is a sample, not a measurement. One run of one prompt is a single data point, results shift when prompts change, and the data reflects one browser at one moment. It is best used alongside a repeated, fixed set of prompts and read as an estimate. Setup assumes comfort with Python 3, Google Chrome or Chromium and the project's dependencies, with installation steps kept in the GitHub README. Compared with hosted AI visibility trackers, it offers no dashboards, scheduling or support for more engines, but it costs nothing and keeps the raw data in the user's own files. Because it automates consumer websites, the terms of those services may not permit it, and the page asks users to check them first.
Free and open source under the MIT licence. Sampling ChatGPT requires your own ChatGPT account, whose cost is separate and not covered on the page.
Learn moreInstallation steps are in the README on the GitHub repository. No other support channel is described, and the tool comes without warranty.
Works with Perplexity and ChatGPT through the browser, uses Selenium with Google Chrome or Chromium, and exports to CSV or JSONL. Code is hosted on GitHub.
Runs prompts from a CSV through Perplexity or ChatGPT in a real Chrome browser via Selenium, paces requests, saves answers with cited source addresses to CSV (optional JSONL), and supports a prompt limit for small first runs.