Crawl finds winning organic videos and paid ads across Instagram, TikTok, and the Meta Ad Library, qualifies the strongest references, and turns them into structured research reports.
Thousands of videos → qualified references → recurring patterns → usable creative ideas
- Searches Instagram, TikTok, and Meta Ad Library references
- Evaluates spoken hooks, visual hooks, and on-screen text separately
- Filters for stronger creative references instead of returning a raw scrape
- Produces local HTML research reports with patterns, examples, and proof links
Download the latest Mac version
Crawl v1 supports Apple-chip Macs, Intel Macs, and macOS 13 or newer. This release is not for Windows.
Choose whichever setup format feels easier:
- Watch Part 1: download and set up Crawl
- Watch Part 2: run Crawl and understand the reports
- Open the illustrated Start Guide
After downloading:
- Unzip
CRAWL.zip. - Open the
crawlfolder. - Double-click
run-crawl.command. - Follow the questions shown in the Terminal window.
You will need an Apify account and either a Gemini API key or a Google Cloud project for Vertex.
- You choose a platform, brand, or research direction.
- Crawl gathers candidate videos and ads through your Apify account.
- Gemini or Vertex analyzes the creative using your Google account.
- Crawl saves the finished report locally on your Mac.
Open the extracted Crawl folder in Claude Code or Codex, add the Start Guide, and ask it to diagnose the error and fix Crawl for you.
Crawl runs locally and uses accounts that you connect yourself. Vishnupriya does not receive or manage your Apify, Gemini, or Google Cloud credentials. Requests and source media are processed by Apify and Google under your accounts. Read PRIVACY.md for details.
The code is available in this repository so you can inspect it and adapt it for your own workflow. The ready-to-run Mac download is packaged separately under Releases because it includes a bundled runtime and is too large for the source repository.
Crawl is released under the MIT License.