II.
Benchmark overview
Reference · livebenchmark:gaia
GAIA overview
General AI Assistants benchmark — real-world agent reasoning tasks.
Attributes
displayName
GAIA
homepageUrl
kind
agent-reasoning
targetsKind
AgentVersion
description
General AI Assistants benchmark — real-world agent reasoning tasks.
Outgoing edges
covers1
- skill-area:agentic-loops·SkillAreaAgentic Loops
Incoming edges
bounds_subject1
- scope-boundary:gaia.scope·ScopeBoundary
evaluated_by1
- eval-run:gaia.claude-code.2025·EvalRun
for_benchmark1
- eval-run:gaia.claude-code.2025·EvalRun
scored_against1
- eval-result:gaia.claude-code.001·EvalResult
split_of1
- test-set:gaia-validation·TestSetGAIA validation split