Phi 4 Multimodal vs.SmolVLM2 2.2B
Passed:
Phi 4 Multimodal 49.4% (41/83)
SmolVLM2 2.2B 36.6% (30/82)
Average request time:
Phi 4 Multimodal 15.92s
SmolVLM2 2.2B 4.16s
Passed:
Phi 4 Multimodal 49.4% (41/83)
SmolVLM2 2.2B 36.6% (30/82)
Average request time:
Phi 4 Multimodal 15.92s
SmolVLM2 2.2B 4.16s
Summary
Annotation Understanding
Phi 4 Multimodal
1 / 2
SmolVLM2 2.2B
2 / 2
CAPTCHA
Phi 4 Multimodal
0 / 2
SmolVLM2 2.2B
0 / 2
Color Identification
Phi 4 Multimodal
0 / 1
SmolVLM2 2.2B
0 / 1
Defect Detection
Phi 4 Multimodal
11 / 15
SmolVLM2 2.2B
9 / 15
Document Understanding
Phi 4 Multimodal
7 / 10
SmolVLM2 2.2B
3 / 10
Localization
Phi 4 Multimodal
0 / 1
SmolVLM2 2.2B
0 / 1
OCR
Phi 4 Multimodal
5 / 9
SmolVLM2 2.2B
2 / 9
Object Counting
Phi 4 Multimodal
0 / 11
SmolVLM2 2.2B
1 / 11
Object Detection
Phi 4 Multimodal
2 / 2
SmolVLM2 2.2B
1 / 2
Object Measurement
Phi 4 Multimodal
0 / 1
SmolVLM2 2.2B
0 / 1
Object Understanding
Phi 4 Multimodal
5 / 11
SmolVLM2 2.2B
3 / 11
Receipt Reading
Phi 4 Multimodal
1 / 1
SmolVLM2 2.2B
0 / 1
Sign Understanding
Phi 4 Multimodal
2 / 2
SmolVLM2 2.2B
2 / 2
Spatial Relations
Phi 4 Multimodal
7 / 17
SmolVLM2 2.2B
10 / 17
Web Action Understanding
Phi 4 Multimodal
2 / 4
SmolVLM2 2.2B
0 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!