# Changelog of News & Article Extractor (`automation-lab/news-article-extractor`) Actor

- **URL**: https://apify.com/automation-lab/news-article-extractor/changelog.md
- **Full Actor documentation**: https://apify.com/automation-lab/news-article-extractor.md

## Changelog

### 2026-09-20

*Reliability*

- Large multi-site jobs now share work fairly across sources, save each result as it completes, and stop claiming new articles shortly before the platform deadline so partial batches finish cleanly instead of timing out.
- Slow HTTP attempts and retry waits are capped by the remaining run budget, preserving already extracted rows and charges.
- Section, listing, and navigation pages are now reported as failed candidates instead of successful articles, so they are never billed as extracted articles.

### 2026-09-10

*Performance*

- Full-article jobs now use safe bounded parallel extraction, reducing runtime for multi-article feeds while preserving per-site limits, result fields, and pay-per-success charging.
- Added `maxConcurrency` (default 2, range 1–10) so users can tune request pressure for slower or rate-limited sources.

*Documentation*

- Expanded the dataset overview to show article content, Markdown, extraction status, and errors.
- Clarified that date filters do not deduplicate articles across separate runs.

### 2026-06-23

*Reliability*

- Fixed task and API inputs that supply start URLs as request-list objects.

### 2026-06-12

*Features*

- Added Markdown article output, character counts, and token estimates for LLM and RAG workflows.

### 2026-04-07

*Features*

- Initial release for discovering and extracting news and blog articles from direct URLs, RSS feeds, and sitemaps.
- Exports clean article text and metadata for analysis, archiving, and content pipelines.