GPT-4.1 vs.Gemma 4 31B
Passed:
GPT-4.1 70.7% (58/82)
Gemma 4 31B 72.2% (39/54)
Average request time:
GPT-4.1 12.46s
Gemma 4 31B 15.30s
Passed:
GPT-4.1 70.7% (58/82)
Gemma 4 31B 72.2% (39/54)
Average request time:
GPT-4.1 12.46s
Gemma 4 31B 15.30s
Summary
Annotation Understanding
GPT-4.1
1 / 2
Gemma 4 31B
0 / 1
CAPTCHA
GPT-4.1
0 / 2
Gemma 4 31B
2 / 2
Color Identification
GPT-4.1
0 / 1
Gemma 4 31B
0 / 0
Defect Detection
GPT-4.1
12 / 15
Gemma 4 31B
6 / 8
Document Understanding
GPT-4.1
7 / 10
Gemma 4 31B
7 / 7
Localization
GPT-4.1
1 / 1
Gemma 4 31B
0 / 1
OCR
GPT-4.1
9 / 9
Gemma 4 31B
4 / 4
Object Counting
GPT-4.1
2 / 11
Gemma 4 31B
2 / 8
Object Detection
GPT-4.1
2 / 2
Gemma 4 31B
1 / 2
Object Measurement
GPT-4.1
0 / 1
Gemma 4 31B
0 / 0
Object Understanding
GPT-4.1
7 / 11
Gemma 4 31B
4 / 7
Receipt Reading
GPT-4.1
1 / 1
Gemma 4 31B
1 / 1
Sign Understanding
GPT-4.1
2 / 2
Gemma 4 31B
1 / 1
Spatial Relations
GPT-4.1
15 / 17
Gemma 4 31B
10 / 13
Web Action Understanding
GPT-4.1
4 / 4
Gemma 4 31B
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!