Multi Turn Base

BFCL · 50 models

Rank	Model	Score
1	xLAM-2-70b-fc-r (FC)	82.5
2	xLAM-2-32b-fc-r (FC)	81.5
3	Claude-Opus-4-5-20251101 (FC)	81
4	xLAM-2-8b-fc-r (FC)	76
5	BitAgent-Bounty-8B	75
6	GLM-4.6 (FC thinking)	74.5
7	xLAM-2-3b-fc-r (FC)	71.5
8	Grok-4-1-fast-reasoning (FC)	70.5
9	Claude-Sonnet-4-5-20250929 (FC)	69
9	Gemini-3-Pro-Preview (FC)	69
11	o3-2025-04-16 (Prompt)	68
12	Arch-Agent-32B	64.5
12	Gemini-3-Pro-Preview (Prompt)	64.5
14	Claude-Haiku-4-5-20251001 (FC)	63.5
15	Moonshotai-Kimi-K2-Instruct (FC)	62
16	Command A Reasoning (FC)	61.5
17	Nanbeige4-3B-Thinking-2511 (FC)	58.5
18	Grok-4-1-fast-non-reasoning (FC)	58
19	Qwen3-235B-A22B-Instruct-2507 (FC)	57.5
20	Nanbeige3.5-Pro-Thinking (FC)	56
20	Qwen3-32B (FC)	56
22	Grok-4-0709 (Prompt)	55.5
23	DeepSeek-V3.2-Exp (Prompt + Thinking)	55
24	GPT-5.2-2025-12-11 (Prompt)	54.5
25	Qwen3-235B-A22B-Instruct-2507 (Prompt)	54
25	Qwen3-32B (Prompt)	54
27	o4-mini-2025-04-16 (FC)	51
28	Qwen3-8B (FC)	50.5
29	ToolACE-2-8B (FC)	49
30	GPT-4.1-2025-04-14 (FC)	47.5
31	xLAM-2-1b-fc-r (FC)	45.5
32	GPT-5-nano-2025-08-07 (FC)	44
32	Grok-4-0709 (FC)	44
34	GPT-4.1-mini-2025-04-14 (FC)	43.5
34	Qwen3-30B-A3B-Instruct-2507 (FC)	43.5
36	Arch-Agent-3B	42
37	DeepSeek-V3.2-Exp (FC)	41.5
37	Gemini-2.5-Flash (FC)	41.5
37	Qwen3-8B (Prompt)	41.5
40	GPT-4.1-nano-2025-04-14 (FC)	39.5
41	Qwen3-14B (FC)	39
42	Command A (FC)	38
43	GPT-5-mini-2025-08-07 (FC)	36.5
43	GPT-5.2-2025-12-11 (FC)	36.5
45	Arch-Agent-1.5B	35.5
46	Qwen3-30B-A3B-Instruct-2507 (Prompt)	33
47	Llama-4-Maverick-17B-128E-Instruct-FP8 (FC)	27
48	Qwen3-4B-Instruct-2507 (FC)	26.5
49	Llama-3.3-70B-Instruct (FC)	26
50	Qwen3-4B-Instruct-2507 (Prompt)	24.5