Gemini 2.5 Flash Preview vs.Claude 4 Sonnet
Passed:
Gemini 2.5 Flash Preview 67.1% (55/82)
Claude 4 Sonnet 70.7% (58/82)
Average request time:
Gemini 2.5 Flash Preview 28.05s
Claude 4 Sonnet 13.67s
Passed:
Gemini 2.5 Flash Preview 67.1% (55/82)
Claude 4 Sonnet 70.7% (58/82)
Average request time:
Gemini 2.5 Flash Preview 28.05s
Claude 4 Sonnet 13.67s
Summary
Annotation Understanding
Gemini 2.5 Flash Preview
2 / 2
Claude 4 Sonnet
1 / 2
CAPTCHA
Gemini 2.5 Flash Preview
2 / 2
Claude 4 Sonnet
2 / 2
Color Identification
Gemini 2.5 Flash Preview
1 / 1
Claude 4 Sonnet
1 / 1
Counting nodes
Gemini 2.5 Flash Preview
0 / 1
Claude 4 Sonnet
0 / 1
Defect Detection
Gemini 2.5 Flash Preview
11 / 15
Claude 4 Sonnet
12 / 15
Document Understanding
Gemini 2.5 Flash Preview
5 / 10
Claude 4 Sonnet
6 / 10
Localization
Gemini 2.5 Flash Preview
0 / 1
Claude 4 Sonnet
0 / 1
OCR
Gemini 2.5 Flash Preview
8 / 9
Claude 4 Sonnet
7 / 9
Object Counting
Gemini 2.5 Flash Preview
3 / 10
Claude 4 Sonnet
3 / 10
Object Detection
Gemini 2.5 Flash Preview
2 / 2
Claude 4 Sonnet
2 / 2
Object Measurement
Gemini 2.5 Flash Preview
0 / 2
Claude 4 Sonnet
1 / 2
Object Understanding
Gemini 2.5 Flash Preview
9 / 10
Claude 4 Sonnet
8 / 10
Receipt Reading
Gemini 2.5 Flash Preview
1 / 1
Claude 4 Sonnet
1 / 1
Sign Understanding
Gemini 2.5 Flash Preview
2 / 2
Claude 4 Sonnet
2 / 2
Spatial Relations
Gemini 2.5 Flash Preview
11 / 17
Claude 4 Sonnet
14 / 17
Web Action Understanding
Gemini 2.5 Flash Preview
4 / 4
Claude 4 Sonnet
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!