Model

gemini-2.5-flash-preview-05-20-23K

Google

deepmind.google

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
Aider Polyglot55.1percent_correctN/AIndependentmediumaider.chat

Abstract Reasoning

Abstract Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ARC-AGI-22.5scoreN/AIndependentlowepoch.ai

Reproduced by the community

Loading community runs…