Skip to content
Longterm Wiki
benchmark-result

Gemini on ARC-AGI-2: 77.1

Child of ARC-AGI-2

Metadata

Source Tablebenchmark_results
Source IDtNdG07wWqh
ParentARC-AGI-2
Children
CreatedApr 24, 2026, 6:59 PM
UpdatedApr 24, 2026, 6:59 PM
SyncedApr 24, 2026, 6:59 PM

Record Data

idtNdG07wWqh
benchmarkIdV5ocpwgWQj
modelIdGemini(ai-model)
score77.1
unitpercent
date2026-02-19
sourceUrl
notesGemini 3.1 Pro on abstract visual reasoning benchmark

Source Check Verdicts

confirmed98% confidence

Last checked: 4/24/2026

Inline sourcing: confirmed

Debug info

Thing ID: tNdG07wWqh

Source Table: benchmark_results

Source ID: tNdG07wWqh

Parent Thing ID: V5ocpwgWQj