Fixed 500 errors on some crawled websites and during server capacity issues, fixed v1 crawl status for pages over 10mb, fixed screenshot undefined response, pushed infrastructure improvements to reduce speed times when scraper fails.
Interact with pages before extracting data to unlock more data from every site. Firecrawl allows performing various actions on web pages before scraping, useful for dynamic content, navigation, and user interaction. Actions include wait, click, write, press, and screenshot operations. Node SDK 1.5.x supports type-safe actions, available in REST API and Python SDK.
Typesafe LLM Extract: E2E type safety for LLM extract in Node SDK 1.5.x, 10x cheaper at 5 credits per extract (down from 50), improved speed and reliability.
Rust SDK v1.0.0: Rust SDK v1 released.
Map Improved Limits: Map smart results limits increased from 100 to 1000.
Faster scrape: Scrape speed improved by 200-600ms depending on website.
General: Numerous infrastructure and API improvements launched. Changelog introduced for future releases.
Fixed an issue where custom header params were not properly being set in v1 API.
Firecrawl V1 introduces a more reliable and developer-friendly API with the following features:
Learn more about it in the documentation.