#3030th of 33 in Data and scraping for AI
llm-scraper
TypeScript library that extracts structured data from web pages with LLMs
- Stars
- 7k
- License
- MIT
- Last commit
- Jun 2026
- Language
- TypeScript
6,190 on 4 Apr 20266,950 on 31 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
LLM Scraper loads a page through Playwright and asks an LLM to fill a Zod or JSON Schema, returning typed objects. It runs on the Vercel AI SDK, so OpenAI, Anthropic, Google, Groq and Ollama models plug in. It can also stream partial objects or generate a reusable Playwright script for a schema.
Who it is for: TypeScript developers extracting structured data from web pages
Strengths
- Schemas in Zod or JSON Schema give type-safe output in TypeScript
- Six input formats: html, raw_html, markdown, text, image, custom
- Code generation produces a reusable Playwright script, avoiding an LLM call per page
- Works with hosted providers and local models via Ollama
Weaknesses
- Library only: no CLI, server or UI
- Needs Playwright and a separately installed AI SDK provider package
- README gives no cost, token-limit or accuracy guidance for large pages
- No tagged release published
What it needs
- no GPU
- Needs Playwright, Zod, Vercel AI SDK provider package, LLM provider API or Ollama
- Models: OpenAI GPT, Anthropic Claude, Google Gemini, Groq Llama, Ollama
Quick answers
Is llm-scraper open source?
Yes. Its code is on GitHub under the MIT license.
Does llm-scraper need a GPU?
No, it runs without one.
Also in Data and scraping for AI
See all 33| Rank | Project | Score |
|---|---|---|
| 1 | mineruDocument parser that converts PDFs, images and Office files to Markdown and JSON | 94 out of 100 |
| 2 | doclingPython library and CLI that converts PDFs, Office files and more into structured text | 91 out of 100 |
| 3 | opendataloader-pdfPDF parser that outputs Markdown, JSON with bounding boxes, and Tagged PDF | 87 out of 100 |
| #4 | markitdownPython library and CLI that converts office files, PDFs and audio to Markdown | 86 out of 100 |
| #5 | agent-reachCLI that gives coding agents read access to login-gated and Chinese web platforms | 85 out of 100 |