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

Tool Use on BFCL V4

76.5Accuracy

Claude-Opus-4.5

29.4441.657553.87566.0925Feb 13, 2026Mar 7, 2026Mar 30, 2026Apr 22, 2026May 15, 2026Jun 7, 2026Jun 30, 2026
Updated 23d ago

Evaluation Results

MethodLinks
76.5---
73.4---
72.9---
72.9---
71---
65.9---
65---
57.9---
2026.02
56.5---
2026.02
53.8---
2026.02
50.51---
2026.02
48.6---
2026.02
47.9---
2026.05
45.6---
2026.02
44.87---
2026.05
42.9---
2026.05
42.7---
2026.05
38.8---
2026.06
36.6235.9856.44.12
2026.06
36.4336.3155.224.88
2026.06
36.3536.0655.224.88
2026.06
36.2735.9455.294.63
2026.05
35.9---
2026.06
35.635.7554.563.38
2026.06
35.5835.7154.034.25
2026.06
34.5735.1352.194
2026.06
34.0233.9852.632.63
2026.06
33.9333.4852.782.75
2026.06
33.8234.2951.962.5
2026.06
33.5333.152.192.63
2026.06
33.4133.9851.222.5
2026.06
32.5532.8150.262.25
2026.06
31.2530.9248.782.13