Claude 3.5 Haiku vs.Gemini 3 Flash (Tools)
Passed:
Claude 3.5 Haiku 65.1% (54/83)
Gemini 3 Flash (Tools) 82.9% (68/82)
Average request time:
Claude 3.5 Haiku 16.75s
Gemini 3 Flash (Tools) 68.14s
Passed:
Claude 3.5 Haiku 65.1% (54/83)
Gemini 3 Flash (Tools) 82.9% (68/82)
Average request time:
Claude 3.5 Haiku 16.75s
Gemini 3 Flash (Tools) 68.14s
Summary
Annotation Understanding
Claude 3.5 Haiku
2 / 2
Gemini 3 Flash (Tools)
0 / 2
CAPTCHA
Claude 3.5 Haiku
0 / 2
Gemini 3 Flash (Tools)
2 / 2
Color Identification
Claude 3.5 Haiku
0 / 1
Gemini 3 Flash (Tools)
0 / 1
Defect Detection
Claude 3.5 Haiku
10 / 15
Gemini 3 Flash (Tools)
14 / 15
Document Understanding
Claude 3.5 Haiku
8 / 10
Gemini 3 Flash (Tools)
9 / 10
Localization
Claude 3.5 Haiku
0 / 1
Gemini 3 Flash (Tools)
1 / 1
OCR
Claude 3.5 Haiku
7 / 9
Gemini 3 Flash (Tools)
9 / 9
Object Counting
Claude 3.5 Haiku
5 / 11
Gemini 3 Flash (Tools)
7 / 11
Object Detection
Claude 3.5 Haiku
2 / 2
Gemini 3 Flash (Tools)
2 / 2
Object Measurement
Claude 3.5 Haiku
0 / 1
Gemini 3 Flash (Tools)
0 / 1
Object Understanding
Claude 3.5 Haiku
7 / 11
Gemini 3 Flash (Tools)
8 / 11
Receipt Reading
Claude 3.5 Haiku
1 / 1
Gemini 3 Flash (Tools)
1 / 1
Sign Understanding
Claude 3.5 Haiku
2 / 2
Gemini 3 Flash (Tools)
2 / 2
Spatial Relations
Claude 3.5 Haiku
8 / 17
Gemini 3 Flash (Tools)
15 / 17
Web Action Understanding
Claude 3.5 Haiku
4 / 4
Gemini 3 Flash (Tools)
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!