BFCL
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
BFCL V3
79.3Overall Accuracy
104
BFCL Multi-Turn v3
78.7Overall Accuracy
69
BFCL
0False Negative Rate
56
BFCL multi-turn V3
68Average Success Rate
48
BFCL
94Accuracy
45
BFCL Task
76Factuality Score
42
BFCL V4
76.5Accuracy
33
BFCL V4
62.3Multi-Turn Success Rate
32
BFCL Individual Tools per Problem
95Execution Accuracy
30
BFCL
83.27Success Rate (Simple)
29
BFCL v4
68.8Score
25
BFCL (Live)
88.25Simple Accuracy
24
BFCL Multi-Turn Base v3
69Greedy Success Rate
24
BFCL Multi-turn
54.75Accuracy
24
BFCL v2
5.31MAT Score
22
BFCL Multi-turn
42.3Accuracy
22
BFCL Single-turn
84.2Accuracy
22
BFCL
77BFCL v4 Score
20
BFCL MT v4
48Success Rate
20
BFCL parallel parallel-multiple Actions
82.2Accuracy
20
BFCL Memory
28.22Task Accuracy
20
BFCL Multi-Parallel v2
87.5Accuracy
20
BFCL Parallel v2
87.5Accuracy
20
BFCL Multi-Parallel v1
90.5Accuracy
20
BFCL Parallel v1
95.5Accuracy
20