← BFCL leaderboards

Non-Live Parallel Multiple AST

BFCL · 50 models

RankModelScore
1Claude-Opus-4-5-20251101 (Prompt)92.5
2Claude-Sonnet-4-5-20250929 (FC)92
2Gemini-2.5-Flash-Lite (FC)92
2Gemini-3-Pro-Preview (Prompt)92
2Qwen3-14B (FC)92
2Qwen3-14B (Prompt)92
7Mistral-Small-2506 (Prompt)91.5
7Open-Mistral-Nemo-2407 (Prompt)91.5
7Qwen3-32B (FC)91.5
7Qwen3-32B (Prompt)91.5
11Llama-3.3-70B-Instruct (FC)91
11Qwen3-235B-A22B-Instruct-2507 (Prompt)91
11ToolACE-2-8B (FC)91
11xLAM-2-32b-fc-r (FC)91
15Grok-4-1-fast-non-reasoning (FC)90.5
15Llama-4-Scout-17B-16E-Instruct (FC)90.5
17Grok-4-1-fast-reasoning (FC)90
17Qwen3-4B-Instruct-2507 (FC)90
19GLM-4.6 (FC thinking)89.5
19Gemini-2.5-Flash-Lite (Prompt)89.5
19Gemma-3-27b-it (Prompt)89.5
19Qwen3-4B-Instruct-2507 (Prompt)89.5
19Qwen3-8B (Prompt)89.5
19palmyra-x-004 (FC)89.5
19xLAM-2-70b-fc-r (FC)89.5
26Arch-Agent-32B89
26Falcon3-7B-Instruct (FC)89
26GPT-4.1-2025-04-14 (Prompt)89
26Qwen3-8B (FC)89
30Claude-Haiku-4-5-20251001 (FC)88.5
30Claude-Opus-4-5-20251101 (FC)88.5
30CoALM-8B88.5
30Command A Reasoning (FC)88.5
30Llama-4-Maverick-17B-128E-Instruct-FP8 (FC)88.5
30Qwen3-30B-A3B-Instruct-2507 (FC)88.5
36Command A (FC)88
36Qwen3-30B-A3B-Instruct-2507 (Prompt)88
38Falcon3-10B-Instruct (FC)87.5
38Gemini-2.5-Flash (Prompt)87.5
40Grok-4-0709 (FC)87
41DeepSeek-V3.2-Exp (Prompt + Thinking)86.5
42Amazon-Nova-2-Lite-v1:0 (FC)86
42Hammer2.1-7b (FC)86
44Mistral-Medium-250585.5
45Nanbeige4-3B-Thinking-2511 (FC)84.5
45Open-Mistral-Nemo-2407 (FC)84.5
47Amazon-Nova-Pro-v1:0 (FC)84
48GPT-4.1-mini-2025-04-14 (Prompt)83.5
48mistral-large-2411 (FC)83.5
48xLAM-2-8b-fc-r (FC)83.5