Function Calling
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
79.3Overall Accuracy
104
Jun 30, 2026
78.7Overall Accuracy
69
May 14, 2026
0False Negative Rate
56
Apr 24, 2026
77.66F1 Score
40
Apr 30, 2026
62.3Multi-Turn Success Rate
32
Jun 8, 2026
95Execution Accuracy
30
Feb 26, 2026
75.95Pass Rate
30
May 29, 2026
83.27Success Rate (Simple)
29
Apr 30, 2026
95.94Exact Match (EM)
28
Jun 23, 2026
41.8Base Score
28
Feb 26, 2026
0.876AST Simple
26
Feb 26, 2026
81.4AST Simple
26
Feb 26, 2026
68.8Score
25
May 21, 2026
88.25Simple Accuracy
24
May 29, 2026
42.3Accuracy
22
Mar 6, 2026
84.2Accuracy
22
Mar 6, 2026
79.17Level-1 Score
20
Jun 16, 2026
28.22Task Accuracy
20
May 19, 2026
69.44Non-live Accuracy
20
Apr 23, 2026
78Atom Score
20
Apr 23, 2026
0.938Accuracy
20
Mar 4, 2026
621.13Input Tokens
19
Feb 26, 2026
64.1FSA
18
Feb 26, 2026
60.5FSA
18
Feb 26, 2026
89FSA
18
Feb 26, 2026