Mistral Small 3.1 24B vs.Cosmos Reason2 8B
Passed:
Mistral Small 3.1 24B 44.6% (37/83)
Cosmos Reason2 8B 63.4% (52/82)
Average request time:
Mistral Small 3.1 24B 15.16s
Cosmos Reason2 8B 2.72s
Passed:
Mistral Small 3.1 24B 44.6% (37/83)
Cosmos Reason2 8B 63.4% (52/82)
Average request time:
Mistral Small 3.1 24B 15.16s
Cosmos Reason2 8B 2.72s
Summary
Annotation Understanding
Mistral Small 3.1 24B
0 / 2
Cosmos Reason2 8B
2 / 2
CAPTCHA
Mistral Small 3.1 24B
2 / 2
Cosmos Reason2 8B
2 / 2
Color Identification
Mistral Small 3.1 24B
1 / 1
Cosmos Reason2 8B
1 / 1
Defect Detection
Mistral Small 3.1 24B
8 / 15
Cosmos Reason2 8B
10 / 15
Document Understanding
Mistral Small 3.1 24B
5 / 10
Cosmos Reason2 8B
5 / 10
Localization
Mistral Small 3.1 24B
0 / 1
Cosmos Reason2 8B
1 / 1
OCR
Mistral Small 3.1 24B
7 / 9
Cosmos Reason2 8B
8 / 9
Object Counting
Mistral Small 3.1 24B
0 / 11
Cosmos Reason2 8B
3 / 11
Object Detection
Mistral Small 3.1 24B
1 / 2
Cosmos Reason2 8B
0 / 2
Object Measurement
Mistral Small 3.1 24B
0 / 1
Cosmos Reason2 8B
0 / 1
Object Understanding
Mistral Small 3.1 24B
4 / 11
Cosmos Reason2 8B
8 / 11
Receipt Reading
Mistral Small 3.1 24B
1 / 1
Cosmos Reason2 8B
1 / 1
Sign Understanding
Mistral Small 3.1 24B
0 / 2
Cosmos Reason2 8B
2 / 2
Spatial Relations
Mistral Small 3.1 24B
6 / 17
Cosmos Reason2 8B
10 / 17
Web Action Understanding
Mistral Small 3.1 24B
4 / 4
Cosmos Reason2 8B
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!