GPT-4.1 vs.Cosmos Reason2 8B
Passed:
GPT-4.1 70.7% (58/82)
Cosmos Reason2 8B 63.4% (52/82)
Average request time:
GPT-4.1 12.46s
Cosmos Reason2 8B 2.72s
Passed:
GPT-4.1 70.7% (58/82)
Cosmos Reason2 8B 63.4% (52/82)
Average request time:
GPT-4.1 12.46s
Cosmos Reason2 8B 2.72s
Summary
Annotation Understanding
GPT-4.1
1 / 2
Cosmos Reason2 8B
2 / 2
CAPTCHA
GPT-4.1
0 / 2
Cosmos Reason2 8B
2 / 2
Color Identification
GPT-4.1
0 / 1
Cosmos Reason2 8B
1 / 1
Defect Detection
GPT-4.1
12 / 15
Cosmos Reason2 8B
10 / 15
Document Understanding
GPT-4.1
7 / 10
Cosmos Reason2 8B
5 / 10
Localization
GPT-4.1
1 / 1
Cosmos Reason2 8B
1 / 1
OCR
GPT-4.1
9 / 9
Cosmos Reason2 8B
8 / 9
Object Counting
GPT-4.1
2 / 11
Cosmos Reason2 8B
3 / 11
Object Detection
GPT-4.1
2 / 2
Cosmos Reason2 8B
0 / 2
Object Measurement
GPT-4.1
0 / 1
Cosmos Reason2 8B
0 / 1
Object Understanding
GPT-4.1
7 / 11
Cosmos Reason2 8B
8 / 11
Receipt Reading
GPT-4.1
1 / 1
Cosmos Reason2 8B
1 / 1
Sign Understanding
GPT-4.1
2 / 2
Cosmos Reason2 8B
2 / 2
Spatial Relations
GPT-4.1
15 / 17
Cosmos Reason2 8B
10 / 17
Web Action Understanding
GPT-4.1
4 / 4
Cosmos Reason2 8B
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!