LegalBench ↗82.49 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · LegalBench · Accuracy ↗
1 more reported result
82.63 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · LegalBench · Accuracy · Vals AI · LegalBench ↗
Google / MODEL EVIDENCE
Gemini 2.5 Flash Preview (9/25)
13 benchmarks · 28 observations · 14 sources
Scores at a glance
13 benchmarks · large dot: Gemini 2.5 Flash Preview (9/25); small dots: closest models where reported
CorpFin v2 ↗58.97 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · CorpFin · Accuracy ↗
1 more reported result
59.75 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · CorpFin · Accuracy · Vals AI · CorpFin ↗
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MortgageTax · Accuracy ↗
3 more reported results
72.73 % · Gemini 2.5 Flash Preview (9/25) (Nonthinking) · TaxEval · Accuracy · Vals AI · TaxEval ↗
61.92 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MortgageTax · Accuracy · Vals AI · MortgageTax ↗
72.4 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · TaxEval · Accuracy · Vals AI · TaxEval ↗
MedCode ↗40.54 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MedCode · Accuracy ↗
1 more reported result
40.33 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MedCode · Accuracy · Vals AI · MedCode ↗
MedScribe ↗77.95 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MedScribe · Accuracy ↗
1 more reported result
78.5 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MedScribe · Accuracy · Vals AI · MedScribe ↗
MedQA (USMLE) ↗91.43 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MedQA · Accuracy ↗
1 more reported result
91.17 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MedQA · Accuracy · Vals AI · MedQA ↗
AIME 2024–2025 · Vals ↗49.79 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · AIME · Accuracy ↗
1 more reported result
51.46 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · AIME · Accuracy · Vals AI · AIME ↗
MGSM (Multilingual GSM) ↗89.85 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MGSM · Accuracy ↗
1 more reported result
89.82 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MGSM · Accuracy · Vals AI · MGSM ↗
GPQA Diamond ↗81.57 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · GPQA Diamond · Accuracy ↗
1 more reported result
76.52 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · GPQA Diamond · Accuracy · Vals AI · GPQA Diamond ↗
MMLU-Pro ↗83.69 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MMLU Pro · Accuracy ↗
1 more reported result
83.66 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MMLU Pro · Accuracy · Vals AI · MMLU Pro ↗
MMMU-Pro ↗79.48 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · MMMU Pro · Accuracy ↗
1 more reported result
80.75 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · MMMU Pro · Accuracy · Vals AI · MMMU Pro ↗
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · SAGE · Accuracy ↗
1 more reported result
39.38 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · SAGE · Accuracy · Vals AI · SAGE ↗
LiveCodeBench ↗75.06 %
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · LiveCodeBench · Accuracy ↗
1 more reported result
76.21 % · Gemini 2.5 Flash Preview (9/25) (Thinking) · LiveCodeBench · Accuracy · Vals AI · LiveCodeBench ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
There are not enough shared scored settings to identify close comparisons reliably.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- AIME — evaluation methodology and results ↗
- CorpFin — evaluation methodology and results ↗
- GPQA Diamond — evaluation methodology and results ↗
- LiveCodeBench — evaluation methodology and results ↗
- LegalBench — evaluation methodology and results ↗
- MedCode — evaluation methodology and results ↗
- MedQA — evaluation methodology and results ↗
- MedScribe — evaluation methodology and results ↗
- MGSM — evaluation methodology and results ↗
- MMLU Pro — evaluation methodology and results ↗
- MMMU Pro — evaluation methodology and results ↗
- MortgageTax — evaluation methodology and results ↗
- SAGE — evaluation methodology and results ↗
- TaxEval — evaluation methodology and results ↗
Names used in source tables
Gemini 2.5 Flash Preview (9/25) (Nonthinking) · Gemini 2.5 Flash Preview (9/25) (Thinking)