ScrapeOpsAI Scraper Builder - Product Updates
Here you will find the latest updates to the ScrapeOps AI Scraper Builder. This page tracks new features, supported languages/frameworks, data schemas, and improvements.
The AI Scraper Builder is now open for Beta users! Sign up for the Beta program to get 20 free scraper generations.
What is the AI Scraper Builder?
The AI Scraper Builder generates production-ready web scraper code from URLs. Simply paste a few sample URLs, and our AI:
- Fetches & analyzes the page structure
- Maps CSS/XPath selectors to your chosen data schema
- Generates complete scraper code in your preferred language and framework
No setup, no guesswork, just clean, working code in minutes.
Supported Page Types
| Page Type | Status |
|---|---|
| Product Details | ✅ Available |
| Product Search | ✅ Available |
| Product Categories | ✅ Available |
| Hotels, real estate, courses, cars, blogs, news, jobs, business directories | Beta |
| Product crawler (URL discovery from listing pages) | Beta |
| Any other page | Dynamic schema |
Beta types are auto-detected and generate a working scraper. Accuracy is strongest on the three product types. A page that matches none of them gets a schema built from that page.
Supported Programming Languages
| Language | Framework | Status |
|---|---|---|
| Python | BeautifulSoup + Requests | ✅ Available |
| Python | Scrapy | ✅ Available |
| Python | Selenium | ✅ Available |
| Python | Playwright | ✅ Available |
| Node.js | Cheerio + Axios | ✅ Available |
| Node.js | Puppeteer | ✅ Available |
| Node.js | Playwright | ✅ Available |
| Go | Colly | 🗓️ Roadmap |
| Java | — | 🗓️ Roadmap |
Changelog
Here are the latest product updates.
- Automatic Self-Healing (Aug '26)
- Claude Code Plugin (May '26)
- Dynamic Schema Generation (Apr '26)
- Beta Page Types (Mar '26)
- Beta Program Now Open (Jan '26)
Automatic Self-Healing (Aug '26)
- New Feature: Generated parsers are checked against the data they are expected to return
- Summary: If coverage drops, ScrapeOps generates a replacement parser and compares it with the current one. The replacement goes live only when it extracts more of the expected data. If a replacement later does worse, the previous parser is restored.
📖 Docs: Automatic Self-Healing
Claude Code Plugin (May '26)
- New Feature: AI Scraper Builder inside Claude Code
- Summary: Generate a scraper, generate a crawler plus scraper, or fix a local parser without leaving the editor. The plugin can write Python, JavaScript, PHP, Ruby, Go, Rust, Java, and C#. Generations use the same account quota as the dashboard.
📖 Docs: Claude Code Plugin
Dynamic Schema Generation (Apr '26)
- New Feature: Custom extraction schema for pages that do not match a known type
- Summary: When auto-detection cannot classify a URL, the builder builds a schema from the page HTML and captured XHR or fetch responses, then runs the same generation pipeline. Forums, portfolios, event listings, and other unmatched pages can produce a scraper.
📖 Docs: Dynamic Schema Generation
Beta Page Types (Mar '26)
- New Feature: Auto-detected page types beyond ecommerce products
- Summary: The builder now classifies hotels, real estate, online courses, cars, blogs, news, jobs, and business directories, plus a product-crawler type that collects detail URLs from listing pages. These types generate end to end and are still being refined, so they are marked beta.
📖 Docs: Supported Page Types
Beta Program Now Open (Jan '26)
- Launch: AI Scraper Builder Beta is now open to users
- New Feature: AI-powered page structure analysis
- New Feature: Automatic CSS/XPath selector mapping
- New Feature: Multi-language code generation (Python, Node.js)
- New Feature: Multi-framework support (Scrapy, Playwright, Puppeteer, Selenium, BeautifulSoup, Cheerio)
- New Feature: Pre-defined data schemas for Product Details, Product Search, and Product Categories
- New Feature: Built-in ScrapeOps Proxy Aggregator integration
- Free Tier: 20 free scraper generations—no credit card required
Get Started
Ready to generate your first AI scraper?