← BFCL leaderboards

Latency 95th Percentile (s)

BFCL · 50 models

RankModelScore
1Qwen3-0.6B (FC)0.96
2Open-Mistral-Nemo-2407 (Prompt)1.32
3Open-Mistral-Nemo-2407 (FC)1.39
4Gemini-2.5-Flash-Lite (Prompt)1.4
5Gemini-2.5-Flash-Lite (FC)1.67
6Amazon-Nova-Micro-v1:0 (FC)1.79
7GPT-4.1-nano-2025-04-14 (Prompt)1.88
8Mistral-Small-2506 (Prompt)2.02
9GPT-4.1-nano-2025-04-14 (FC)2.26
10Qwen3-14B (Prompt)2.3
11GPT-4.1-mini-2025-04-14 (FC)2.4
11Llama-3.1-Nemotron-Ultra-253B-v1 (FC)2.4
13Mistral-small-2506 (FC)2.5
14GPT-4.1-2025-04-14 (Prompt)2.53
15Command R7B (FC)2.69
16Qwen3-30B-A3B-Instruct-2507 (Prompt)2.84
17Mistral-Medium-25052.86
18Claude-Haiku-4-5-20251001 (FC)3.15
19Amazon-Nova-Pro-v1:0 (FC)3.29
20GPT-4.1-mini-2025-04-14 (Prompt)3.38
21Granite-4.0-350m (FC)3.44
22Claude-Haiku-4-5-20251001 (Prompt)3.77
23GPT-4.1-2025-04-14 (FC)4.01
24mistral-large-2411 (Prompt)4.08
25Mistral-Medium-2505 (FC)4.19
26mistral-large-2411 (FC)4.68
27Command A (FC)4.94
28GPT-5.2-2025-12-11 (FC)5.26
29Arch-Agent-1.5B5.3
30Grok-4-1-fast-non-reasoning (FC)5.34
31Claude-Opus-4-5-20251101 (Prompt)5.52
32Gemini-2.5-Flash (FC)5.62
33Gemini-2.5-Flash (Prompt)6.09
34Qwen3-235B-A22B-Instruct-2507 (FC)6.27
35xLAM-2-1b-fc-r (FC)6.52
36Claude-Sonnet-4-5-20250929 (Prompt)6.66
37Claude-Sonnet-4-5-20250929 (FC)7.27
38Claude-Opus-4-5-20251101 (FC)7.56
39Qwen3-235B-A22B-Instruct-2507 (Prompt)7.61
40palmyra-x-004 (FC)8.04
41Arch-Agent-3B8.19
42Command A Reasoning (FC)8.39
42o3-2025-04-16 (FC)8.39
44xLAM-2-3b-fc-r (FC)8.79
45o4-mini-2025-04-16 (FC)9.33
46Llama-3.2-1B-Instruct (FC)9.77
47Hammer2.1-0.5b (FC)9.86
48Granite-20b-FunctionCalling (FC)9.97
49o4-mini-2025-04-16 (Prompt)10.19
50Qwen3-0.6B (Prompt)10.31