Qwen 3.5 Plus vs.GPT-5.4
Passed:
Qwen 3.5 Plus 68.3% (56/82)
GPT-5.4 79.3% (65/82)
Average request time:
Qwen 3.5 Plus 4.37s
GPT-5.4 9.40s
Passed:
Qwen 3.5 Plus 68.3% (56/82)
GPT-5.4 79.3% (65/82)
Average request time:
Qwen 3.5 Plus 4.37s
GPT-5.4 9.40s
Summary
Annotation Understanding
Qwen 3.5 Plus
0 / 2
GPT-5.4
2 / 2
CAPTCHA
Qwen 3.5 Plus
2 / 2
GPT-5.4
0 / 2
Color Identification
Qwen 3.5 Plus
1 / 1
GPT-5.4
1 / 1
Defect Detection
Qwen 3.5 Plus
10 / 15
GPT-5.4
13 / 15
Document Understanding
Qwen 3.5 Plus
9 / 10
GPT-5.4
9 / 10
Localization
Qwen 3.5 Plus
1 / 1
GPT-5.4
1 / 1
OCR
Qwen 3.5 Plus
7 / 9
GPT-5.4
9 / 9
Object Counting
Qwen 3.5 Plus
3 / 11
GPT-5.4
4 / 11
Object Detection
Qwen 3.5 Plus
0 / 2
GPT-5.4
2 / 2
Object Measurement
Qwen 3.5 Plus
1 / 1
GPT-5.4
0 / 1
Object Understanding
Qwen 3.5 Plus
9 / 11
GPT-5.4
8 / 11
Receipt Reading
Qwen 3.5 Plus
1 / 1
GPT-5.4
1 / 1
Sign Understanding
Qwen 3.5 Plus
2 / 2
GPT-5.4
2 / 2
Spatial Relations
Qwen 3.5 Plus
11 / 17
GPT-5.4
15 / 17
Web Action Understanding
Qwen 3.5 Plus
4 / 4
GPT-5.4
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!