Gemini 2.5 Deep Think Benchmark Details

Gemini 2.5 Deep Think currently shows benchmark results led by LiveCodeBench (12 / 127, score 87.60), AIME2025 (11 / 106, score 99.20), FrontierMath (12 / 60, score 29).

Benchmark Results

Gemini 2.5 Deep Think

Benchmark Results

General Knowledge

1 evaluations

Benchmark / mode

Score

Rank/total

HLE

Deep

34.80

82 / 185

Coding and Software Engineer

1 evaluations

Benchmark / mode

Score

Rank/total

LiveCodeBench

Deep

87.60

12 / 127

Math and Reasoning

4 evaluations

Benchmark / mode

Score

Rank/total

AIME2025

Deep

99.20

11 / 106

IMO-ProofBench

Thinking Enabled

37.60

5 / 16

FrontierMath

Deep

12 / 60

FrontierMath - Tier 4

Standard Mode

10.40

32 / 80

Compare with other models