news-crawler-LM: Small Long-Context Model for HTML-to-JSON Extraction
July 24, 2026
news-crawler-LM is a small language model fine-tuned to convert raw HTML into structured JSON and plaintext. It automates the extraction of headlines, authors, and article bodies, outperforming standard baselines in HTML-to-Markdown tasks while reducing the computational overhead of larger models.
HOW THIS AFFECTS YOU
●
builderYou can use this to build more efficient, low-latency web scraping pipelines without the cost of frontier models.