DataLearner logo

Gemini 2.5 Flash-Lite-Preview-09-2025 Benchmark Details

Gemini 2.5 Flash-Lite-Preview-09-2025 currently shows benchmark results led by MMMU (41 / 74, score 72.70), DocVQA (4 / 5, score 92), MedScribe (54 / 66, score 75.82).

Benchmark Results

Gemini 2.5 Flash-Lite-Preview-09-2025

Benchmark Results

Thinking
Tool usage

Visual Understanding

1 evaluations
Benchmark / mode
Score
Rank/total
MMMU
Standard Mode
72.70
41 / 74

Documents & Charts

1 evaluations
Benchmark / mode
Score
Rank/total
DocVQA
Standard Mode
92
4 / 5

Cross-capability Suites

1 evaluations
Benchmark / mode
Score
Rank/total
LiveBench
Thinking Level · High
42.39
110 / 117

Clinical Workflows

4 evaluations
Benchmark / mode
Score
Rank/total
MedScribe
Standard ModeTools
75.82
54 / 66
MedScribe
Thinking ModeTools
66.88
64 / 66
MedCode
Standard ModeTools
27.08
64 / 64
MedCode
Thinking ModeTools
34.19
53 / 64

Service Workflows

2 evaluations
Benchmark / mode
Score
Rank/total
SAGE
Standard ModeTools
27.97
62 / 64
SAGE
Thinking ModeTools
30.78
56 / 64