Claude 3.5 Haiku vs.Phi 4 Multimodal
Passed:
Claude 3.5 Haiku 65.1% (54/83)
Phi 4 Multimodal 49.4% (41/83)
Average request time:
Claude 3.5 Haiku 16.75s
Phi 4 Multimodal 15.92s
Passed:
Claude 3.5 Haiku 65.1% (54/83)
Phi 4 Multimodal 49.4% (41/83)
Average request time:
Claude 3.5 Haiku 16.75s
Phi 4 Multimodal 15.92s
Summary
Annotation Understanding
Claude 3.5 Haiku
2 / 2
Phi 4 Multimodal
1 / 2
CAPTCHA
Claude 3.5 Haiku
0 / 2
Phi 4 Multimodal
0 / 2
Color Identification
Claude 3.5 Haiku
0 / 1
Phi 4 Multimodal
0 / 1
Counting nodes
Claude 3.5 Haiku
0 / 1
Phi 4 Multimodal
0 / 1
Defect Detection
Claude 3.5 Haiku
10 / 15
Phi 4 Multimodal
11 / 15
Document Understanding
Claude 3.5 Haiku
8 / 10
Phi 4 Multimodal
7 / 10
Localization
Claude 3.5 Haiku
0 / 1
Phi 4 Multimodal
0 / 1
OCR
Claude 3.5 Haiku
7 / 9
Phi 4 Multimodal
5 / 9
Object Counting
Claude 3.5 Haiku
5 / 10
Phi 4 Multimodal
0 / 10
Object Detection
Claude 3.5 Haiku
2 / 2
Phi 4 Multimodal
2 / 2
Object Measurement
Claude 3.5 Haiku
0 / 2
Phi 4 Multimodal
0 / 2
Object Understanding
Claude 3.5 Haiku
7 / 11
Phi 4 Multimodal
5 / 11
Receipt Reading
Claude 3.5 Haiku
1 / 1
Phi 4 Multimodal
1 / 1
Sign Understanding
Claude 3.5 Haiku
2 / 2
Phi 4 Multimodal
2 / 2
Spatial Relations
Claude 3.5 Haiku
8 / 17
Phi 4 Multimodal
7 / 17
Web Action Understanding
Claude 3.5 Haiku
4 / 4
Phi 4 Multimodal
2 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!