ResearchDatasetsEarth-BenchFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsTool Use and ReasoningEarth-Bench70.82Accuracy62Active Tool ExplorationEarth-Bench39,080Tokens per Question (AP)9Earth Observation Agent Task CompletionEarth-Bench Lite65Spectrum Score7