ChatGPT-4o (Medium Reasoning) vs.Claude 3.5 Haiku
Passed:
ChatGPT-4o (Medium Reasoning) 75.6% (62/82)
Claude 3.5 Haiku 65.1% (54/83)
Average request time:
ChatGPT-4o (Medium Reasoning) 20.48s
Claude 3.5 Haiku 16.75s
Passed:
ChatGPT-4o (Medium Reasoning) 75.6% (62/82)
Claude 3.5 Haiku 65.1% (54/83)
Average request time:
ChatGPT-4o (Medium Reasoning) 20.48s
Claude 3.5 Haiku 16.75s
Summary
Annotation Understanding
ChatGPT-4o (Medium Reasoning)
1 / 2
Claude 3.5 Haiku
2 / 2
CAPTCHA
ChatGPT-4o (Medium Reasoning)
2 / 2
Claude 3.5 Haiku
0 / 2
Color Identification
ChatGPT-4o (Medium Reasoning)
0 / 1
Claude 3.5 Haiku
0 / 1
Defect Detection
ChatGPT-4o (Medium Reasoning)
12 / 15
Claude 3.5 Haiku
10 / 15
Document Understanding
ChatGPT-4o (Medium Reasoning)
7 / 10
Claude 3.5 Haiku
8 / 10
Localization
ChatGPT-4o (Medium Reasoning)
1 / 1
Claude 3.5 Haiku
0 / 1
OCR
ChatGPT-4o (Medium Reasoning)
9 / 9
Claude 3.5 Haiku
7 / 9
Object Counting
ChatGPT-4o (Medium Reasoning)
3 / 11
Claude 3.5 Haiku
5 / 11
Object Detection
ChatGPT-4o (Medium Reasoning)
2 / 2
Claude 3.5 Haiku
2 / 2
Object Measurement
ChatGPT-4o (Medium Reasoning)
1 / 1
Claude 3.5 Haiku
0 / 1
Object Understanding
ChatGPT-4o (Medium Reasoning)
8 / 11
Claude 3.5 Haiku
7 / 11
Receipt Reading
ChatGPT-4o (Medium Reasoning)
1 / 1
Claude 3.5 Haiku
1 / 1
Sign Understanding
ChatGPT-4o (Medium Reasoning)
2 / 2
Claude 3.5 Haiku
2 / 2
Spatial Relations
ChatGPT-4o (Medium Reasoning)
15 / 17
Claude 3.5 Haiku
8 / 17
Web Action Understanding
ChatGPT-4o (Medium Reasoning)
4 / 4
Claude 3.5 Haiku
4 / 4
Contribute a Prompt
Have an idea for a prompt? Open a pull request on the project repository!