aiaimodel.cards
← All benchmarks

AIME 2026 (no tools)

Problems from the American Invitational Mathematics Examination, a hard high-school math competition.

  1. 1Gemma 4 31B · Google
    89.2%

How each maker reported it

  • Gemma 4 31B Benchmark results

    Instruction-tuned models, as published in Google's model card. OmniDocBench is edit distance (lower is better). Audio benchmarks are omitted because 31B has no audio input.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.