pypi / firecrawl-py
firecrawl-py
What changed in firecrawl-py 2, from its official release notes, upgrade guides and advisories. Sources checked 2026-10-19.
1 breaking1 security11 new APIs and options2 behavior and defaults1 fixes and other
See which of these apply to your code, ranked, in your agent:
npx @unstale/cli initfirecrawl-py 216 changes
Breaking changes 1
The org flag now uses
disabled/allowed/forcedvalues, and the legacy boolean request shape has been removed.
Security 1
Adds missing negative-limit validation in the Python, Java, and Go SDKs.
New APIs and options 11
Adds an org flag for custom robots user agents independently of
ignoreRobots.Adds a
/v2/searchparameter for excluding specific domains from results.Exposes the Docker harness startup timeout through docker-compose for self-hosted users needing longer startup windows.
Adds a
/scrapeformat that returns matching sentences, code blocks, and table rows from the source page.Adds a
/v2/searchparameter for scoping results to specific domains.Adds a
/scrapeoption that serves results exclusively from Firecrawl's index, with cache misses returning 404.Adds multipart file parsing for uploads up to 50 MB, returning a standard Document with clean, LLM-ready output.
Adds search-result feedback submissions, with accepted submissions refunding a credit subject to a daily cap and age limit.
Adds a
/scrapeformat that returns a grounded answer indata.question.Adds a crawl option for evaluating robots.txt rules and crawl delays against a custom agent string.
Adds a
/scrapeformat that returnsdocument.videoas a signed URL; unsupported URLs raise an error.
Behavior and default changes 2
PDF billing now reflects pages processed instead of raw page count.
Raises the PDF upload size limit from 10 MB to 30 MB.
Fixes and other changes 1
Fixes the native link filter and JS SDK crawl payload so they honor and include the custom robots user agent.