GPT-4.1 vs.GPT-5 (high reasoning)
Passed:
GPT-4.1 72.3% (60/83)
GPT-5 (high reasoning) 73.5% (61/83)
Average request time:
GPT-4.1 9.74s
GPT-5 (high reasoning) 26.96s
Passed:
GPT-4.1 72.3% (60/83)
GPT-5 (high reasoning) 73.5% (61/83)
Average request time:
GPT-4.1 9.74s
GPT-5 (high reasoning) 26.96s
Summary
Annotation Understanding
GPT-4.1
1 / 2
GPT-5 (high reasoning)
1 / 2
CAPTCHA
GPT-4.1
2 / 2
GPT-5 (high reasoning)
0 / 2
Color Identification
GPT-4.1
0 / 1
GPT-5 (high reasoning)
1 / 1
Counting nodes
GPT-4.1
0 / 1
GPT-5 (high reasoning)
0 / 1
Defect Detection
GPT-4.1
12 / 15
GPT-5 (high reasoning)
12 / 15
Document Understanding
GPT-4.1
8 / 10
GPT-5 (high reasoning)
8 / 10
Localization
GPT-4.1
1 / 1
GPT-5 (high reasoning)
1 / 1
OCR
GPT-4.1
9 / 9
GPT-5 (high reasoning)
9 / 9
Object Counting
GPT-4.1
2 / 10
GPT-5 (high reasoning)
3 / 10
Object Detection
GPT-4.1
2 / 2
GPT-5 (high reasoning)
1 / 2
Object Measurement
GPT-4.1
0 / 2
GPT-5 (high reasoning)
0 / 2
Object Understanding
GPT-4.1
8 / 11
GPT-5 (high reasoning)
9 / 11
Receipt Reading
GPT-4.1
1 / 1
GPT-5 (high reasoning)
1 / 1
Sign Understanding
GPT-4.1
2 / 2
GPT-5 (high reasoning)
2 / 2
Spatial Relations
GPT-4.1
14 / 17
GPT-5 (high reasoning)
15 / 17
Web Action Understanding
GPT-4.1
4 / 4
GPT-5 (high reasoning)
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!