II.
Benchmark overview
Reference · livebenchmark:browse-comp
BrowseComp overview
OpenAI benchmark for browsing agents on hard, hard-to-find web research tasks designed to require deep, persistent search rather than surface lookup.
Attributes
displayName
BrowseComp
homepageUrl
kind
web-agent
targetsKind
AgentVersion
description
OpenAI benchmark for browsing agents on hard, hard-to-find web
research tasks designed to require deep, persistent search rather
than surface lookup.
Outgoing edges
applies_to1
- domain:web-development·DomainWeb Development
covers2
- skill-area:deep-web-research·SkillAreaDeep Web Research
- skill-area:browser-automation·SkillAreaBrowser Automation
Incoming edges
None.