Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Function Calling on BFCL Live v3
Loading...
77.9
Overall Accuracy
Qwen3-4B
48.78
56.34
63.9
71.46
Feb 4, 2026
Overall Accuracy
Function Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Overall Accuracy
Function Accuracy
Qwen3-4B
Speedup ratio=1×
2026.02
77.9
94.5
Qwen2.5-14B
Speedup ratio=1×
2026.02
76.6
95.9
ST-Qwen3-4B
Speedup ratio=3.9×
2026.02
76.4
96.3
ST-Qwen2.5-14B
Speedup ratio=4.4×
2026.02
75.9
97.1
ST-Qwen2.5-7B
Speedup ratio=4.5×
2026.02
75.8
96.7
Qwen2.5-7B
Speedup ratio=1×
2026.02
75.4
95.2
Qwen2.5-3B
Speedup ratio=1×
2026.02
73
94.3
ST-Qwen2.5-3B
Speedup ratio=3.8×
2026.02
68
95.4
Qwen2.5-1.5B
Speedup ratio=1×
2026.02
67.4
90.4
ST-Qwen2.5-1.5B
Speedup ratio=3.5×
2026.02
63.6
94.3
ST-Qwen2.5-0.5B
Speedup ratio=3.1×
2026.02
57.2
91.6
Qwen2.5-0.5B
Speedup ratio=1×
2026.02
49.9
86.4
Feedback
Search any
task
Search any
task