raznem/parsera

Health v2 · maintenance only

Adoption not part of the score

1350 stars · 78 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Parsera is a lightweight Python library for scraping websites using LLMs, letting users define elements to extract with natural-language descriptions and get structured JSON results. It runs on Playwright for page fetching, supports custom LangChain models (including local Ollama/HuggingFace models), proxies, cookies, scrolling, custom Playwright scripts, and offers a CLI and Docker usage.

Use cases

• extract structured data from web pages with an llm
• scrape tables from websites into
• scrape pages that require login or javascript rendering
• run llm-based scraping with a local ollama model
• extract fields from a webpage by describing them in plain english
• scrape infinite-scroll pages
• scrape websites through a proxy

When to choose

• you want simple, schema-driven web extraction without writing CSS selectors
• target pages are dynamic or JavaScript-rendered and need a real browser
• you want to plug in any LangChain chat model, including local ones
• you need typed/structured output from unstructured pages

When to avoid

• you need high-volume, low-cost scraping where deterministic selectors or BeautifulSoup suffice
• you cannot send page content to an LLM API due to privacy or cost constraints
• you need a fully managed scalable scraping service rather than a local library

Facets

library · maturity active

web-scraping nlp llm-inference parser data-science crawlers developer-tools python cli cross-platform llm-scraping playwright structured-extraction browser-automation data-extraction natural-language-processing automation docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="raznem/parsera")