Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Judge Evaluation on Magis-Bench

7.79Score

Gemini-3 Pro

3.45324.57915.7056.8309Mar 10, 2026
Updated 2mo ago

Evaluation Results

MethodLinks
2026.03
7.79
2026.03
7.48
2026.03
6.99
2026.03
6.66
2026.03
5.55
2026.03
5.08
2026.03
4.97
2026.03
4.88
2026.03
4.52
2026.03
4.5
2026.03
4.49
2026.03
4.47
2026.03
4.25
2026.03
3.67
2026.03
3.62