OpenAI O4 Mini (Medium Reasoning)

75.9% (63/83)

Passed

14.26s

Average Request Time

Summary

Ranked #1 on leaderboard for Web Action Understanding, Receipt Reading, Localization, Sign Understanding, Color Identification, Annotation Understanding, CAPTCHA.

• Web Action Understanding → 4 / 4
• Receipt Reading → 1 / 1
• Localization → 1 / 1
• Sign Understanding → 2 / 2
• Color Identification → 1 / 1
• Annotation Understanding → 2 / 2
• CAPTCHA → 2 / 2
• Defect Detection → 14 / 15
• OCR → 8 / 9
• Spatial Relations → 15 / 17
• Object Understanding → 9 / 11
• Document Understanding → 8 / 10
• Object Detection → 1 / 2
• Object Counting → 2 / 10
• Object Measurement → 0 / 2
• Counting nodes → 0 / 1

Contribute a Prompt

Have an idea for a prompt? Open a pull request on the project repository!

Compare with another model