HN Showcase
Back to browse
LucidExtractor – Extract web data in plain English, no selectors

LucidExtractor – Extract web data in plain English, no selectors

by yukendiran_j·Feb 26, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●MidCrowd Pleaser

AI-powered selectors sound good, but Firecrawl, JinaAI, and Bright Data already do this—for less friction.

Strengths
  • •No-selector UX (plain English prompt) lowers friction for non-programmers compared to XPath/CSS selector tools.
  • •Generous free tier (500 credits, no card required) and clear pricing removes adoption friction for trials.
Weaknesses
  • •LLM-based extraction is table stakes now—Firecrawl, JinaAI, and commercial scrapers offer identical core value proposition.
  • •Gemini 2.5 Flash dependence means you're competing on API latency and cost, not differentiation; no mention of success rate, hallucination handling, or real-world benchmarks.
Target Audience

Data analysts, market researchers, SEO professionals, non-technical users doing one-off scraping

Similar To

Firecrawl · JinaAI · Bright Data

Post Description

I built a web scraping tool where you describe the data you want in plain English instead of writing CSS selectors or XPath.

The problem: Traditional scrapers break when websites change their HTML structure. You spend more time maintaining selectors than actually using the data.

How it works: Send a URL + an AI prompt like "extract all product names and prices as JSON" and the AI reads the page like a human, returning structured data.

Tech stack: FastAPI backend, Gemini 2.5 Flash for extraction, Playwright for rendering, deployed on Google Cloud Run.

Free tier: 500 credits, no credit card required. Would love feedback from HN on the approach and pricing.

Similar Projects