GPT-5 (high reasoning) vs.GPT-5.2
Passed:
GPT-5 (high reasoning) 73.5% (61/83)
GPT-5.2 78.0% (64/82)
Average request time:
GPT-5 (high reasoning) 26.96s
GPT-5.2 6.94s
Passed:
GPT-5 (high reasoning) 73.5% (61/83)
GPT-5.2 78.0% (64/82)
Average request time:
GPT-5 (high reasoning) 26.96s
GPT-5.2 6.94s
Summary
Annotation Understanding
GPT-5 (high reasoning)
1 / 2
GPT-5.2
1 / 2
CAPTCHA
GPT-5 (high reasoning)
0 / 2
GPT-5.2
0 / 2
Color Identification
GPT-5 (high reasoning)
1 / 1
GPT-5.2
0 / 1
Defect Detection
GPT-5 (high reasoning)
12 / 15
GPT-5.2
14 / 15
Document Understanding
GPT-5 (high reasoning)
8 / 10
GPT-5.2
9 / 10
Localization
GPT-5 (high reasoning)
1 / 1
GPT-5.2
1 / 1
OCR
GPT-5 (high reasoning)
9 / 9
GPT-5.2
9 / 9
Object Counting
GPT-5 (high reasoning)
3 / 11
GPT-5.2
3 / 11
Object Detection
GPT-5 (high reasoning)
1 / 2
GPT-5.2
2 / 2
Object Measurement
GPT-5 (high reasoning)
0 / 1
GPT-5.2
0 / 1
Object Understanding
GPT-5 (high reasoning)
9 / 11
GPT-5.2
8 / 11
Receipt Reading
GPT-5 (high reasoning)
1 / 1
GPT-5.2
1 / 1
Sign Understanding
GPT-5 (high reasoning)
2 / 2
GPT-5.2
2 / 2
Spatial Relations
GPT-5 (high reasoning)
15 / 17
GPT-5.2
15 / 17
Web Action Understanding
GPT-5 (high reasoning)
4 / 4
GPT-5.2
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!