Author
Konstantin Lopukhin
Head of R&D
Konstantin is Zyte’s Head of R&D, where he leads research and development into machine learning methods to improve extraction, parsing, and content quality. He has also participated in Kaggle competitions, achieving a Grandmaster Title, and contributes to the community with talks, sharing code, and knowledge.
Claude Fable 5 is the new best model for writing scrapers
We ran nine models, including the new GPT-5.6, through our Zyte Scraping Code Benchmark. Claude Fable 5 puts Sol and others in the shade - but it’s pricey, and the best extraction code still depends on the best infrastructure.
Claude Sonnet 4.6 is the new best model for writing scrapers
Claude Sonnet 4.6 is now the top model in Zyte’s Web Scraping Copilot benchmark, narrowly beating Gemini 3 Pro on extraction quality, with a small increase in code complexity.
Gemini 3.0 Pro is the new best model for writing scrapers
Gemini 3.0 Pro outperforms GPT-5, Claude, and other leading LLMs in Zyte’s Web Scraping Copilot benchmarks, delivering the highest code accuracy and lowest complexity. See full results, pros, cons, and recommendations for production workflows.
Agentic web scraping: Hype, reality and what happens next
From broken parsers to context limits, today’s AI agents have real challenges—but with the right tools and orchestration, they could reshape how we extract web data.
The Fly, The Parrot & The Thinking Machine: The Rise of Reasoning LLMs
By leveraging the power of LLMs to reason about web page structures and data relationships, we can automate tasks that previously required significant human intervention.
Article data extraction | How to Maximize Quality
Learn how different tools are used to maximize the quality of your news and article data extraction. Understand why it's important and how to scale extraction.
Article extraction: Maximizing data quality