Gemma 4 31B vs.Kimi k2.5
Passed:
Gemma 4 31B 72.2% (39/54)
Kimi k2.5 73.5% (50/68)
Average request time:
Gemma 4 31B 15.30s
Kimi k2.5 44.70s
Passed:
Gemma 4 31B 72.2% (39/54)
Kimi k2.5 73.5% (50/68)
Average request time:
Gemma 4 31B 15.30s
Kimi k2.5 44.70s
Summary
Annotation Understanding
Gemma 4 31B
0 / 1
Kimi k2.5
1 / 1
CAPTCHA
Gemma 4 31B
2 / 2
Kimi k2.5
2 / 2
Color Identification
Gemma 4 31B
0 / 0
Kimi k2.5
1 / 1
Defect Detection
Gemma 4 31B
6 / 8
Kimi k2.5
8 / 14
Document Understanding
Gemma 4 31B
7 / 7
Kimi k2.5
8 / 8
Localization
Gemma 4 31B
0 / 1
Kimi k2.5
1 / 1
OCR
Gemma 4 31B
4 / 4
Kimi k2.5
7 / 9
Object Counting
Gemma 4 31B
2 / 8
Kimi k2.5
1 / 3
Object Detection
Gemma 4 31B
1 / 2
Kimi k2.5
1 / 2
Object Measurement
Gemma 4 31B
0 / 0
Kimi k2.5
0 / 1
Object Understanding
Gemma 4 31B
4 / 7
Kimi k2.5
9 / 11
Receipt Reading
Gemma 4 31B
1 / 1
Kimi k2.5
1 / 1
Sign Understanding
Gemma 4 31B
1 / 1
Kimi k2.5
2 / 2
Spatial Relations
Gemma 4 31B
10 / 13
Kimi k2.5
12 / 15
Web Action Understanding
Gemma 4 31B
4 / 4
Kimi k2.5
3 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!